Please enter the commit message for your changes. Lines starting
with '#' will be ignored, and an empty message aborts the commit. On branch main Your branch is up to date with 'origin/main'. Changes to be committed: new file: .claude/skills/algorithmic-art/.openskills.json new file: .claude/skills/algorithmic-art/LICENSE.txt new file: .claude/skills/algorithmic-art/SKILL.md new file: .claude/skills/algorithmic-art/templates/generator_template.js new file: .claude/skills/algorithmic-art/templates/viewer.html new file: .claude/skills/brand-guidelines/.openskills.json new file: .claude/skills/brand-guidelines/LICENSE.txt new file: .claude/skills/brand-guidelines/SKILL.md new file: .claude/skills/canvas-design/.openskills.json new file: .claude/skills/canvas-design/LICENSE.txt new file: .claude/skills/canvas-design/SKILL.md new file: .claude/skills/canvas-design/canvas-fonts/ArsenalSC-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/ArsenalSC-Regular.ttf new file: .claude/skills/canvas-design/canvas-fonts/BigShoulders-Bold.ttf new file: .claude/skills/canvas-design/canvas-fonts/BigShoulders-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/BigShoulders-Regular.ttf new file: .claude/skills/canvas-design/canvas-fonts/Boldonse-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/Boldonse-Regular.ttf new file: .claude/skills/canvas-design/canvas-fonts/BricolageGrotesque-Bold.ttf new file: .claude/skills/canvas-design/canvas-fonts/BricolageGrotesque-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/BricolageGrotesque-Regular.ttf new file: .claude/skills/canvas-design/canvas-fonts/CrimsonPro-Bold.ttf new file: .claude/skills/canvas-design/canvas-fonts/CrimsonPro-Italic.ttf new file: .claude/skills/canvas-design/canvas-fonts/CrimsonPro-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/CrimsonPro-Regular.ttf new file: .claude/skills/canvas-design/canvas-fonts/DMMono-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/DMMono-Regular.ttf new file: .claude/skills/canvas-design/canvas-fonts/EricaOne-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/EricaOne-Regular.ttf new file: .claude/skills/canvas-design/canvas-fonts/GeistMono-Bold.ttf new file: .claude/skills/canvas-design/canvas-fonts/GeistMono-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/GeistMono-Regular.ttf new file: .claude/skills/canvas-design/canvas-fonts/Gloock-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/Gloock-Regular.ttf new file: .claude/skills/canvas-design/canvas-fonts/IBMPlexMono-Bold.ttf new file: .claude/skills/canvas-design/canvas-fonts/IBMPlexMono-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/IBMPlexMono-Regular.ttf new file: .claude/skills/canvas-design/canvas-fonts/IBMPlexSerif-Bold.ttf new file: .claude/skills/canvas-design/canvas-fonts/IBMPlexSerif-BoldItalic.ttf new file: .claude/skills/canvas-design/canvas-fonts/IBMPlexSerif-Italic.ttf new file: .claude/skills/canvas-design/canvas-fonts/IBMPlexSerif-Regular.ttf new file: .claude/skills/canvas-design/canvas-fonts/InstrumentSans-Bold.ttf new file: .claude/skills/canvas-design/canvas-fonts/InstrumentSans-BoldItalic.ttf new file: .claude/skills/canvas-design/canvas-fonts/InstrumentSans-Italic.ttf new file: .claude/skills/canvas-design/canvas-fonts/InstrumentSans-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/InstrumentSans-Regular.ttf new file: .claude/skills/canvas-design/canvas-fonts/InstrumentSerif-Italic.ttf new file: .claude/skills/canvas-design/canvas-fonts/InstrumentSerif-Regular.ttf new file: .claude/skills/canvas-design/canvas-fonts/Italiana-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/Italiana-Regular.ttf new file: .claude/skills/canvas-design/canvas-fonts/JetBrainsMono-Bold.ttf new file: .claude/skills/canvas-design/canvas-fonts/JetBrainsMono-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/JetBrainsMono-Regular.ttf new file: .claude/skills/canvas-design/canvas-fonts/Jura-Light.ttf new file: .claude/skills/canvas-design/canvas-fonts/Jura-Medium.ttf new file: .claude/skills/canvas-design/canvas-fonts/Jura-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/LibreBaskerville-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/LibreBaskerville-Regular.ttf new file: .claude/skills/canvas-design/canvas-fonts/Lora-Bold.ttf new file: .claude/skills/canvas-design/canvas-fonts/Lora-BoldItalic.ttf new file: .claude/skills/canvas-design/canvas-fonts/Lora-Italic.ttf new file: .claude/skills/canvas-design/canvas-fonts/Lora-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/Lora-Regular.ttf new file: .claude/skills/canvas-design/canvas-fonts/NationalPark-Bold.ttf new file: .claude/skills/canvas-design/canvas-fonts/NationalPark-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/NationalPark-Regular.ttf new file: .claude/skills/canvas-design/canvas-fonts/NothingYouCouldDo-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/NothingYouCouldDo-Regular.ttf new file: .claude/skills/canvas-design/canvas-fonts/Outfit-Bold.ttf new file: .claude/skills/canvas-design/canvas-fonts/Outfit-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/Outfit-Regular.ttf new file: .claude/skills/canvas-design/canvas-fonts/PixelifySans-Medium.ttf new file: .claude/skills/canvas-design/canvas-fonts/PixelifySans-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/PoiretOne-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/PoiretOne-Regular.ttf new file: .claude/skills/canvas-design/canvas-fonts/RedHatMono-Bold.ttf new file: .claude/skills/canvas-design/canvas-fonts/RedHatMono-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/RedHatMono-Regular.ttf new file: .claude/skills/canvas-design/canvas-fonts/Silkscreen-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/Silkscreen-Regular.ttf new file: .claude/skills/canvas-design/canvas-fonts/SmoochSans-Medium.ttf new file: .claude/skills/canvas-design/canvas-fonts/SmoochSans-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/Tektur-Medium.ttf new file: .claude/skills/canvas-design/canvas-fonts/Tektur-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/Tektur-Regular.ttf new file: .claude/skills/canvas-design/canvas-fonts/WorkSans-Bold.ttf new file: .claude/skills/canvas-design/canvas-fonts/WorkSans-BoldItalic.ttf new file: .claude/skills/canvas-design/canvas-fonts/WorkSans-Italic.ttf new file: .claude/skills/canvas-design/canvas-fonts/WorkSans-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/WorkSans-Regular.ttf new file: .claude/skills/canvas-design/canvas-fonts/YoungSerif-OFL.txt new file: .claude/skills/canvas-design/canvas-fonts/YoungSerif-Regular.ttf new file: .claude/skills/doc-coauthoring/.openskills.json new file: .claude/skills/doc-coauthoring/SKILL.md new file: .claude/skills/docx/.openskills.json new file: .claude/skills/docx/LICENSE.txt new file: .claude/skills/docx/SKILL.md new file: .claude/skills/docx/scripts/__init__.py new file: .claude/skills/docx/scripts/accept_changes.py new file: .claude/skills/docx/scripts/comment.py new file: .claude/skills/docx/scripts/office/helpers/__init__.py new file: .claude/skills/docx/scripts/office/helpers/merge_runs.py new file: .claude/skills/docx/scripts/office/helpers/simplify_redlines.py new file: .claude/skills/docx/scripts/office/pack.py new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/dml-chart.xsd new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/dml-chartDrawing.xsd new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/dml-diagram.xsd new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/dml-lockedCanvas.xsd new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/dml-main.xsd new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/dml-picture.xsd new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/dml-spreadsheetDrawing.xsd new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/dml-wordprocessingDrawing.xsd new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/pml.xsd new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-additionalCharacteristics.xsd new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-bibliography.xsd new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-commonSimpleTypes.xsd new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-customXmlDataProperties.xsd new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-customXmlSchemaProperties.xsd new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-documentPropertiesCustom.xsd new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-documentPropertiesExtended.xsd new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-documentPropertiesVariantTypes.xsd new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-math.xsd new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-relationshipReference.xsd new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/sml.xsd new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/vml-main.xsd new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/vml-officeDrawing.xsd new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/vml-presentationDrawing.xsd new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/vml-spreadsheetDrawing.xsd new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/vml-wordprocessingDrawing.xsd new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/wml.xsd new file: .claude/skills/docx/scripts/office/schemas/ISO-IEC29500-4_2016/xml.xsd new file: .claude/skills/docx/scripts/office/schemas/ecma/fouth-edition/opc-contentTypes.xsd new file: .claude/skills/docx/scripts/office/schemas/ecma/fouth-edition/opc-coreProperties.xsd new file: .claude/skills/docx/scripts/office/schemas/ecma/fouth-edition/opc-digSig.xsd new file: .claude/skills/docx/scripts/office/schemas/ecma/fouth-edition/opc-relationships.xsd new file: .claude/skills/docx/scripts/office/schemas/mce/mc.xsd new file: .claude/skills/docx/scripts/office/schemas/microsoft/wml-2010.xsd new file: .claude/skills/docx/scripts/office/schemas/microsoft/wml-2012.xsd new file: .claude/skills/docx/scripts/office/schemas/microsoft/wml-2018.xsd new file: .claude/skills/docx/scripts/office/schemas/microsoft/wml-cex-2018.xsd new file: .claude/skills/docx/scripts/office/schemas/microsoft/wml-cid-2016.xsd new file: .claude/skills/docx/scripts/office/schemas/microsoft/wml-sdtdatahash-2020.xsd new file: .claude/skills/docx/scripts/office/schemas/microsoft/wml-symex-2015.xsd new file: .claude/skills/docx/scripts/office/soffice.py new file: .claude/skills/docx/scripts/office/unpack.py new file: .claude/skills/docx/scripts/office/validate.py new file: .claude/skills/docx/scripts/office/validators/__init__.py new file: .claude/skills/docx/scripts/office/validators/base.py new file: .claude/skills/docx/scripts/office/validators/docx.py new file: .claude/skills/docx/scripts/office/validators/pptx.py new file: .claude/skills/docx/scripts/office/validators/redlining.py new file: .claude/skills/docx/scripts/templates/comments.xml new file: .claude/skills/docx/scripts/templates/commentsExtended.xml new file: .claude/skills/docx/scripts/templates/commentsExtensible.xml new file: .claude/skills/docx/scripts/templates/commentsIds.xml new file: .claude/skills/docx/scripts/templates/people.xml new file: .claude/skills/frontend-design/.openskills.json new file: .claude/skills/frontend-design/LICENSE.txt new file: .claude/skills/frontend-design/SKILL.md new file: .claude/skills/internal-comms/.openskills.json new file: .claude/skills/internal-comms/LICENSE.txt new file: .claude/skills/internal-comms/SKILL.md new file: .claude/skills/internal-comms/examples/3p-updates.md new file: .claude/skills/internal-comms/examples/company-newsletter.md new file: .claude/skills/internal-comms/examples/faq-answers.md new file: .claude/skills/internal-comms/examples/general-comms.md new file: .claude/skills/mcp-builder/.openskills.json new file: .claude/skills/mcp-builder/LICENSE.txt new file: .claude/skills/mcp-builder/SKILL.md new file: .claude/skills/mcp-builder/reference/evaluation.md new file: .claude/skills/mcp-builder/reference/mcp_best_practices.md new file: .claude/skills/mcp-builder/reference/node_mcp_server.md new file: .claude/skills/mcp-builder/reference/python_mcp_server.md new file: .claude/skills/mcp-builder/scripts/connections.py new file: .claude/skills/mcp-builder/scripts/evaluation.py new file: .claude/skills/mcp-builder/scripts/example_evaluation.xml new file: .claude/skills/mcp-builder/scripts/requirements.txt new file: .claude/skills/pdf/.openskills.json new file: .claude/skills/pdf/LICENSE.txt new file: .claude/skills/pdf/SKILL.md new file: .claude/skills/pdf/forms.md new file: .claude/skills/pdf/reference.md new file: .claude/skills/pdf/scripts/check_bounding_boxes.py new file: .claude/skills/pdf/scripts/check_fillable_fields.py new file: .claude/skills/pdf/scripts/convert_pdf_to_images.py new file: .claude/skills/pdf/scripts/create_validation_image.py new file: .claude/skills/pdf/scripts/extract_form_field_info.py new file: .claude/skills/pdf/scripts/extract_form_structure.py new file: .claude/skills/pdf/scripts/fill_fillable_fields.py new file: .claude/skills/pdf/scripts/fill_pdf_form_with_annotations.py new file: .claude/skills/pptx/.openskills.json new file: .claude/skills/pptx/LICENSE.txt new file: .claude/skills/pptx/SKILL.md new file: .claude/skills/pptx/editing.md new file: .claude/skills/pptx/pptxgenjs.md new file: .claude/skills/pptx/scripts/__init__.py new file: .claude/skills/pptx/scripts/add_slide.py new file: .claude/skills/pptx/scripts/clean.py new file: .claude/skills/pptx/scripts/office/helpers/__init__.py new file: .claude/skills/pptx/scripts/office/helpers/merge_runs.py new file: .claude/skills/pptx/scripts/office/helpers/simplify_redlines.py new file: .claude/skills/pptx/scripts/office/pack.py new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/dml-chart.xsd new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/dml-chartDrawing.xsd new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/dml-diagram.xsd new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/dml-lockedCanvas.xsd new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/dml-main.xsd new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/dml-picture.xsd new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/dml-spreadsheetDrawing.xsd new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/dml-wordprocessingDrawing.xsd new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/pml.xsd new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-additionalCharacteristics.xsd new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-bibliography.xsd new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-commonSimpleTypes.xsd new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-customXmlDataProperties.xsd new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-customXmlSchemaProperties.xsd new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-documentPropertiesCustom.xsd new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-documentPropertiesExtended.xsd new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-documentPropertiesVariantTypes.xsd new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-math.xsd new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-relationshipReference.xsd new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/sml.xsd new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/vml-main.xsd new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/vml-officeDrawing.xsd new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/vml-presentationDrawing.xsd new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/vml-spreadsheetDrawing.xsd new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/vml-wordprocessingDrawing.xsd new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/wml.xsd new file: .claude/skills/pptx/scripts/office/schemas/ISO-IEC29500-4_2016/xml.xsd new file: .claude/skills/pptx/scripts/office/schemas/ecma/fouth-edition/opc-contentTypes.xsd new file: .claude/skills/pptx/scripts/office/schemas/ecma/fouth-edition/opc-coreProperties.xsd new file: .claude/skills/pptx/scripts/office/schemas/ecma/fouth-edition/opc-digSig.xsd new file: .claude/skills/pptx/scripts/office/schemas/ecma/fouth-edition/opc-relationships.xsd new file: .claude/skills/pptx/scripts/office/schemas/mce/mc.xsd new file: .claude/skills/pptx/scripts/office/schemas/microsoft/wml-2010.xsd new file: .claude/skills/pptx/scripts/office/schemas/microsoft/wml-2012.xsd new file: .claude/skills/pptx/scripts/office/schemas/microsoft/wml-2018.xsd new file: .claude/skills/pptx/scripts/office/schemas/microsoft/wml-cex-2018.xsd new file: .claude/skills/pptx/scripts/office/schemas/microsoft/wml-cid-2016.xsd new file: .claude/skills/pptx/scripts/office/schemas/microsoft/wml-sdtdatahash-2020.xsd new file: .claude/skills/pptx/scripts/office/schemas/microsoft/wml-symex-2015.xsd new file: .claude/skills/pptx/scripts/office/soffice.py new file: .claude/skills/pptx/scripts/office/unpack.py new file: .claude/skills/pptx/scripts/office/validate.py new file: .claude/skills/pptx/scripts/office/validators/__init__.py new file: .claude/skills/pptx/scripts/office/validators/base.py new file: .claude/skills/pptx/scripts/office/validators/docx.py new file: .claude/skills/pptx/scripts/office/validators/pptx.py new file: .claude/skills/pptx/scripts/office/validators/redlining.py new file: .claude/skills/pptx/scripts/thumbnail.py new file: .claude/skills/skill-creator/.openskills.json new file: .claude/skills/skill-creator/LICENSE.txt new file: .claude/skills/skill-creator/SKILL.md new file: .claude/skills/skill-creator/agents/analyzer.md new file: .claude/skills/skill-creator/agents/comparator.md new file: .claude/skills/skill-creator/agents/grader.md new file: .claude/skills/skill-creator/assets/eval_review.html new file: .claude/skills/skill-creator/eval-viewer/generate_review.py new file: .claude/skills/skill-creator/eval-viewer/viewer.html new file: .claude/skills/skill-creator/references/schemas.md new file: .claude/skills/skill-creator/scripts/__init__.py new file: .claude/skills/skill-creator/scripts/aggregate_benchmark.py new file: .claude/skills/skill-creator/scripts/generate_report.py new file: .claude/skills/skill-creator/scripts/improve_description.py new file: .claude/skills/skill-creator/scripts/package_skill.py new file: .claude/skills/skill-creator/scripts/quick_validate.py new file: .claude/skills/skill-creator/scripts/run_eval.py new file: .claude/skills/skill-creator/scripts/run_loop.py new file: .claude/skills/skill-creator/scripts/utils.py new file: .claude/skills/slack-gif-creator/.openskills.json new file: .claude/skills/slack-gif-creator/LICENSE.txt new file: .claude/skills/slack-gif-creator/SKILL.md new file: .claude/skills/slack-gif-creator/core/easing.py new file: .claude/skills/slack-gif-creator/core/frame_composer.py new file: .claude/skills/slack-gif-creator/core/gif_builder.py new file: .claude/skills/slack-gif-creator/core/validators.py new file: .claude/skills/slack-gif-creator/requirements.txt new file: .claude/skills/template/.openskills.json new file: .claude/skills/template/SKILL.md new file: .claude/skills/theme-factory/.openskills.json new file: .claude/skills/theme-factory/LICENSE.txt new file: .claude/skills/theme-factory/SKILL.md new file: .claude/skills/theme-factory/theme-showcase.pdf new file: .claude/skills/theme-factory/themes/arctic-frost.md new file: .claude/skills/theme-factory/themes/botanical-garden.md new file: .claude/skills/theme-factory/themes/desert-rose.md new file: .claude/skills/theme-factory/themes/forest-canopy.md new file: .claude/skills/theme-factory/themes/golden-hour.md new file: .claude/skills/theme-factory/themes/midnight-galaxy.md new file: .claude/skills/theme-factory/themes/modern-minimalist.md new file: .claude/skills/theme-factory/themes/ocean-depths.md new file: .claude/skills/theme-factory/themes/sunset-boulevard.md new file: .claude/skills/theme-factory/themes/tech-innovation.md new file: .claude/skills/web-artifacts-builder/.openskills.json new file: .claude/skills/web-artifacts-builder/LICENSE.txt new file: .claude/skills/web-artifacts-builder/SKILL.md new file: .claude/skills/web-artifacts-builder/scripts/bundle-artifact.sh new file: .claude/skills/web-artifacts-builder/scripts/init-artifact.sh new file: .claude/skills/web-artifacts-builder/scripts/shadcn-components.tar.gz new file: .claude/skills/webapp-testing/.openskills.json new file: .claude/skills/webapp-testing/LICENSE.txt new file: .claude/skills/webapp-testing/SKILL.md new file: .claude/skills/webapp-testing/examples/console_logging.py new file: .claude/skills/webapp-testing/examples/element_discovery.py new file: .claude/skills/webapp-testing/examples/static_html_automation.py new file: .claude/skills/webapp-testing/scripts/with_server.py new file: .claude/skills/xlsx/.openskills.json new file: .claude/skills/xlsx/LICENSE.txt new file: .claude/skills/xlsx/SKILL.md new file: .claude/skills/xlsx/scripts/office/helpers/__init__.py new file: .claude/skills/xlsx/scripts/office/helpers/merge_runs.py new file: .claude/skills/xlsx/scripts/office/helpers/simplify_redlines.py new file: .claude/skills/xlsx/scripts/office/pack.py new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/dml-chart.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/dml-chartDrawing.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/dml-diagram.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/dml-lockedCanvas.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/dml-main.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/dml-picture.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/dml-spreadsheetDrawing.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/dml-wordprocessingDrawing.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/pml.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-additionalCharacteristics.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-bibliography.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-commonSimpleTypes.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-customXmlDataProperties.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-customXmlSchemaProperties.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-documentPropertiesCustom.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-documentPropertiesExtended.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-documentPropertiesVariantTypes.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-math.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/shared-relationshipReference.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/sml.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/vml-main.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/vml-officeDrawing.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/vml-presentationDrawing.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/vml-spreadsheetDrawing.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/vml-wordprocessingDrawing.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/wml.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ISO-IEC29500-4_2016/xml.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ecma/fouth-edition/opc-contentTypes.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ecma/fouth-edition/opc-coreProperties.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ecma/fouth-edition/opc-digSig.xsd new file: .claude/skills/xlsx/scripts/office/schemas/ecma/fouth-edition/opc-relationships.xsd new file: .claude/skills/xlsx/scripts/office/schemas/mce/mc.xsd new file: .claude/skills/xlsx/scripts/office/schemas/microsoft/wml-2010.xsd new file: .claude/skills/xlsx/scripts/office/schemas/microsoft/wml-2012.xsd new file: .claude/skills/xlsx/scripts/office/schemas/microsoft/wml-2018.xsd new file: .claude/skills/xlsx/scripts/office/schemas/microsoft/wml-cex-2018.xsd new file: .claude/skills/xlsx/scripts/office/schemas/microsoft/wml-cid-2016.xsd new file: .claude/skills/xlsx/scripts/office/schemas/microsoft/wml-sdtdatahash-2020.xsd new file: .claude/skills/xlsx/scripts/office/schemas/microsoft/wml-symex-2015.xsd new file: .claude/skills/xlsx/scripts/office/soffice.py new file: .claude/skills/xlsx/scripts/office/unpack.py new file: .claude/skills/xlsx/scripts/office/validate.py new file: .claude/skills/xlsx/scripts/office/validators/__init__.py new file: .claude/skills/xlsx/scripts/office/validators/base.py new file: .claude/skills/xlsx/scripts/office/validators/docx.py new file: .claude/skills/xlsx/scripts/office/validators/pptx.py new file: .claude/skills/xlsx/scripts/office/validators/redlining.py new file: .claude/skills/xlsx/scripts/recalc.py new file: .env.example new file: .gitignore new file: config/mcp.json new file: config/models.json new file: config/personalities.json new file: docs/AGENTS.md new file: docs/AI_IMPLEMENTATION.md new file: docs/AI_INTEGRATION_COMPLETE.md new file: docs/AI_QUICKSTART.md new file: docs/AI_SUMMARY.md new file: docs/CHANGELOG.md new file: docs/CONFIG_GUIDE.md new file: docs/FIXES.md new file: docs/PROJECT_REFACTOR.md new file: docs/README.md new file: docs/README_INDEX.md new file: examples/ai_example.py new file: main.py new file: pytest.ini new file: requirements.txt new file: scripts/migrate_to_vector_db.py new file: skills/cmd_zip_skill/README.md new file: skills/cmd_zip_skill/__init__.py new file: skills/cmd_zip_skill/main.py new file: skills/cmd_zip_skill/skill.json new file: skills/cmd_zip_skill_1772465404375/README.md new file: skills/cmd_zip_skill_1772465404375/__init__.py new file: skills/cmd_zip_skill_1772465404375/main.py new file: skills/cmd_zip_skill_1772465404375/skill.json new file: skills/cmd_zip_skill_1772465434774/README.md new file: skills/cmd_zip_skill_1772465434774/__init__.py new file: skills/cmd_zip_skill_1772465434774/main.py new file: skills/cmd_zip_skill_1772465434774/skill.json new file: skills/cmd_zip_skill_1772465467809/README.md new file: skills/cmd_zip_skill_1772465467809/__init__.py new file: skills/cmd_zip_skill_1772465467809/main.py new file: skills/cmd_zip_skill_1772465467809/skill.json new file: skills/cmd_zip_skill_1772465652075/README.md new file: skills/cmd_zip_skill_1772465652075/__init__.py new file: skills/cmd_zip_skill_1772465652075/main.py new file: skills/cmd_zip_skill_1772465652075/skill.json new file: skills/cmd_zip_skill_1772465685352/README.md new file: skills/cmd_zip_skill_1772465685352/__init__.py new file: skills/cmd_zip_skill_1772465685352/main.py new file: skills/cmd_zip_skill_1772465685352/skill.json new file: skills/cmd_zip_skill_1772465936294/README.md new file: skills/cmd_zip_skill_1772465936294/__init__.py new file: skills/cmd_zip_skill_1772465936294/main.py new file: skills/cmd_zip_skill_1772465936294/skill.json new file: skills/cmd_zip_skill_1772465966322/README.md new file: skills/cmd_zip_skill_1772465966322/__init__.py new file: skills/cmd_zip_skill_1772465966322/main.py new file: skills/cmd_zip_skill_1772465966322/skill.json new file: skills/cmd_zip_skill_1772466071278/README.md new file: skills/cmd_zip_skill_1772466071278/__init__.py new file: skills/cmd_zip_skill_1772466071278/main.py new file: skills/cmd_zip_skill_1772466071278/skill.json new file: skills/skills_creator/README.md new file: skills/skills_creator/__init__.py new file: skills/skills_creator/main.py new file: skills/skills_creator/skill.json new file: src/__init__.py new file: src/ai/__init__.py new file: src/ai/base.py new file: src/ai/client.py new file: src/ai/docs/README.md new file: src/ai/mcp/__init__.py new file: src/ai/mcp/base.py new file: src/ai/mcp/servers/__init__.py new file: src/ai/mcp/servers/filesystem.py new file: src/ai/memory.py new file: src/ai/models/__init__.py new file: src/ai/models/anthropic_model.py new file: src/ai/models/openai_model.py new file: src/ai/personality.py new file: src/ai/skills/__init__.py new file: src/ai/skills/base.py new file: src/ai/task_manager.py new file: src/ai/vector_store/__init__.py new file: src/ai/vector_store/base.py new file: src/ai/vector_store/chroma_store.py new file: src/ai/vector_store/json_store.py new file: src/core/__init__.py new file: src/core/bot.py new file: src/core/config.py new file: src/handlers/__init__.py new file: src/handlers/message_handler.py new file: src/handlers/message_handler_ai.py new file: src/utils/__init__.py new file: src/utils/logger.py new file: start.bat new file: tests/test_ai.py
This commit is contained in:
Mimikko-zeus
14
src/ai/__init__.py
Normal file
14
src/ai/__init__.py
Normal file
@@ -0,0 +1,14 @@
|
||||
"""
|
||||
AI模块 - 提供AI模型接入、人格系统、记忆系统和长任务处理能力
|
||||
"""
|
||||
from .client import AIClient
|
||||
from .personality import PersonalitySystem
|
||||
from .memory import MemorySystem
|
||||
from .task_manager import LongTaskManager
|
||||
|
||||
__all__ = [
|
||||
'AIClient',
|
||||
'PersonalitySystem',
|
||||
'MemorySystem',
|
||||
'LongTaskManager'
|
||||
]
|
||||
130
src/ai/base.py
Normal file
130
src/ai/base.py
Normal file
@@ -0,0 +1,130 @@
|
||||
"""
|
||||
AI模型基类和接口定义
|
||||
"""
|
||||
from abc import ABC, abstractmethod
|
||||
from typing import List, Dict, Any, Optional, AsyncIterator
|
||||
from dataclasses import dataclass
|
||||
from enum import Enum
|
||||
|
||||
|
||||
class ModelProvider(Enum):
|
||||
"""AI模型提供商"""
|
||||
OPENAI = "openai"
|
||||
ANTHROPIC = "anthropic"
|
||||
DEEPSEEK = "deepseek"
|
||||
QWEN = "qwen"
|
||||
LOCAL = "local"
|
||||
|
||||
|
||||
@dataclass
|
||||
class Message:
|
||||
"""消息对象"""
|
||||
role: str # system, user, assistant, tool
|
||||
content: str
|
||||
name: Optional[str] = None
|
||||
tool_calls: Optional[List[Dict]] = None
|
||||
tool_call_id: Optional[str] = None
|
||||
|
||||
|
||||
@dataclass
|
||||
class ModelConfig:
|
||||
"""模型配置"""
|
||||
provider: ModelProvider
|
||||
model_name: str
|
||||
api_key: Optional[str] = None
|
||||
api_base: Optional[str] = None
|
||||
temperature: float = 0.7
|
||||
max_tokens: int = 4096
|
||||
top_p: float = 1.0
|
||||
frequency_penalty: float = 0.0
|
||||
presence_penalty: float = 0.0
|
||||
timeout: int = 60
|
||||
stream: bool = False
|
||||
|
||||
|
||||
class BaseAIModel(ABC):
|
||||
"""AI模型基类"""
|
||||
|
||||
def __init__(self, config: ModelConfig):
|
||||
self.config = config
|
||||
|
||||
@abstractmethod
|
||||
async def chat(
|
||||
self,
|
||||
messages: List[Message],
|
||||
tools: Optional[List[Dict]] = None,
|
||||
**kwargs
|
||||
) -> Message:
|
||||
"""同步对话"""
|
||||
pass
|
||||
|
||||
@abstractmethod
|
||||
async def chat_stream(
|
||||
self,
|
||||
messages: List[Message],
|
||||
tools: Optional[List[Dict]] = None,
|
||||
**kwargs
|
||||
) -> AsyncIterator[str]:
|
||||
"""流式对话"""
|
||||
pass
|
||||
|
||||
@abstractmethod
|
||||
async def embed(self, text: str) -> List[float]:
|
||||
"""文本嵌入"""
|
||||
pass
|
||||
|
||||
|
||||
@dataclass
|
||||
class ToolDefinition:
|
||||
"""工具定义"""
|
||||
name: str
|
||||
description: str
|
||||
parameters: Dict[str, Any]
|
||||
function: callable
|
||||
|
||||
|
||||
class ToolRegistry:
|
||||
"""工具注册表"""
|
||||
|
||||
def __init__(self):
|
||||
self._tools: Dict[str, ToolDefinition] = {}
|
||||
|
||||
def register(self, tool: ToolDefinition):
|
||||
"""注册工具"""
|
||||
self._tools[tool.name] = tool
|
||||
|
||||
def unregister(self, name: str) -> bool:
|
||||
"""卸载工具"""
|
||||
if name not in self._tools:
|
||||
return False
|
||||
del self._tools[name]
|
||||
return True
|
||||
|
||||
def unregister_by_prefix(self, prefix: str) -> int:
|
||||
"""按前缀卸载工具"""
|
||||
to_remove = [name for name in self._tools.keys() if name.startswith(prefix)]
|
||||
for name in to_remove:
|
||||
del self._tools[name]
|
||||
return len(to_remove)
|
||||
|
||||
def get(self, name: str) -> Optional[ToolDefinition]:
|
||||
"""获取工具"""
|
||||
return self._tools.get(name)
|
||||
|
||||
def list(self) -> List[ToolDefinition]:
|
||||
"""列出所有工具"""
|
||||
return list(self._tools.values())
|
||||
|
||||
def to_openai_format(self) -> List[Dict]:
|
||||
"""转换为OpenAI工具格式"""
|
||||
return [
|
||||
{
|
||||
"type": "function",
|
||||
"function": {
|
||||
"name": tool.name,
|
||||
"description": tool.description,
|
||||
"parameters": tool.parameters
|
||||
}
|
||||
}
|
||||
for tool in self._tools.values()
|
||||
]
|
||||
475
src/ai/client.py
Normal file
475
src/ai/client.py
Normal file
@@ -0,0 +1,475 @@
|
||||
"""
|
||||
AI瀹㈡埛绔?- 鏁村悎鎵€鏈堿I鍔熻兘
|
||||
"""
|
||||
import inspect
|
||||
import json
|
||||
import re
|
||||
from typing import List, Optional, Dict, Any, AsyncIterator, Tuple
|
||||
from pathlib import Path
|
||||
from .base import ModelConfig, ModelProvider, Message, ToolRegistry
|
||||
from .models import OpenAIModel, AnthropicModel
|
||||
from .personality import PersonalitySystem
|
||||
from .memory import MemorySystem
|
||||
from .task_manager import LongTaskManager
|
||||
from src.utils.logger import setup_logger
|
||||
|
||||
logger = setup_logger('AIClient')
|
||||
|
||||
|
||||
class AIClient:
|
||||
"""AI瀹㈡埛绔?- 缁熶竴鎺ュ彛"""
|
||||
|
||||
def __init__(
|
||||
self,
|
||||
model_config: ModelConfig,
|
||||
embed_config: Optional[ModelConfig] = None,
|
||||
data_dir: Path = Path("data/ai"),
|
||||
use_vector_db: bool = True
|
||||
):
|
||||
self.config = model_config
|
||||
self.data_dir = data_dir
|
||||
self.data_dir.mkdir(parents=True, exist_ok=True)
|
||||
|
||||
# 初始化主模型
|
||||
self.model = self._create_model(model_config)
|
||||
|
||||
# 初始化嵌入模型（如果提供）
|
||||
self.embed_model = None
|
||||
if embed_config:
|
||||
self.embed_model = self._create_model(embed_config)
|
||||
logger.info(
|
||||
f"嵌入模型初始化完成: {embed_config.provider.value}/{embed_config.model_name}"
|
||||
)
|
||||
|
||||
# 初始化工具注册表
|
||||
self.tools = ToolRegistry()
|
||||
|
||||
# 初始化人格系统
|
||||
self.personality = PersonalitySystem(
|
||||
config_path=data_dir / "personalities.json"
|
||||
)
|
||||
|
||||
# 初始化记忆系统
|
||||
self.memory = MemorySystem(
|
||||
storage_path=data_dir / "long_term_memory.json",
|
||||
embed_func=self._embed_wrapper,
|
||||
importance_evaluator=self._evaluate_memory_importance,
|
||||
use_vector_db=use_vector_db
|
||||
)
|
||||
|
||||
# 初始化长任务管理器
|
||||
self.task_manager = LongTaskManager(
|
||||
storage_path=data_dir / "tasks.json"
|
||||
)
|
||||
|
||||
logger.info(
|
||||
f"AI 客户端初始化完成: {model_config.provider.value}/{model_config.model_name}"
|
||||
)
|
||||
|
||||
def _create_model(self, config: ModelConfig):
|
||||
"""创建模型实例。"""
|
||||
if config.provider == ModelProvider.OPENAI:
|
||||
return OpenAIModel(config)
|
||||
elif config.provider == ModelProvider.ANTHROPIC:
|
||||
return AnthropicModel(config)
|
||||
elif config.provider in [ModelProvider.DEEPSEEK, ModelProvider.QWEN]:
|
||||
# DeepSeek 和 Qwen 使用 OpenAI 兼容接口
|
||||
return OpenAIModel(config)
|
||||
else:
|
||||
raise ValueError(f"不支持的模型提供商: {config.provider}")
|
||||
|
||||
async def _embed_wrapper(self, text: str) -> List[float]:
|
||||
"""嵌入向量包装器。"""
|
||||
try:
|
||||
# 如果有独立的嵌入模型，优先使用
|
||||
if self.embed_model:
|
||||
return await self.embed_model.embed(text)
|
||||
# 否则尝试使用主模型
|
||||
return await self.model.embed(text)
|
||||
except NotImplementedError:
|
||||
# 如果都不支持嵌入，返回 None（记忆系统会降级）
|
||||
logger.warning("Current model does not support embeddings; vector retrieval disabled")
|
||||
return None
|
||||
except Exception as e:
|
||||
logger.error(f"生成嵌入向量失败: {e}")
|
||||
return None
|
||||
|
||||
@staticmethod
|
||||
def _parse_importance_score(raw: str) -> float:
|
||||
text = (raw or "").strip()
|
||||
if not text:
|
||||
raise ValueError("empty importance response")
|
||||
|
||||
try:
|
||||
parsed = json.loads(text)
|
||||
if isinstance(parsed, (int, float)):
|
||||
return float(parsed)
|
||||
if isinstance(parsed, dict):
|
||||
for key in ["importance", "score", "value"]:
|
||||
if key in parsed:
|
||||
return float(parsed[key])
|
||||
except Exception:
|
||||
pass
|
||||
|
||||
match = re.search(r"-?\d+(?:\.\d+)?", text)
|
||||
if not match:
|
||||
raise ValueError(f"cannot parse importance score: {text}")
|
||||
return float(match.group(0))
|
||||
|
||||
async def _evaluate_memory_importance(
|
||||
self, content: str, metadata: Optional[Dict] = None
|
||||
) -> float:
|
||||
"""
|
||||
调用主模型评估记忆重要性，返回 [0, 1] 分值。
|
||||
"""
|
||||
system_prompt = (
|
||||
"你是记忆重要性评估器。请根据输入内容判断该信息是否值得长期记忆。"
|
||||
"输出一个 0 到 1 的数字，数字越大表示越重要。"
|
||||
"只输出数字，不要输出任何解释、单位或多余文本。"
|
||||
)
|
||||
payload = json.dumps(
|
||||
{"content": content, "metadata": metadata or {}},
|
||||
ensure_ascii=False,
|
||||
)
|
||||
messages = [
|
||||
Message(role="system", content=system_prompt),
|
||||
Message(role="user", content=payload),
|
||||
]
|
||||
|
||||
try:
|
||||
response = await self.model.chat(
|
||||
messages=messages,
|
||||
tools=None,
|
||||
temperature=0.0,
|
||||
max_tokens=16,
|
||||
)
|
||||
score = self._parse_importance_score(response.content)
|
||||
return max(0.0, min(1.0, score))
|
||||
except Exception as e:
|
||||
logger.warning(f"memory importance evaluation failed, fallback to neutral score: {e}")
|
||||
return 0.5
|
||||
|
||||
async def chat(
|
||||
self,
|
||||
user_id: str,
|
||||
user_message: str,
|
||||
system_prompt: Optional[str] = None,
|
||||
use_memory: bool = True,
|
||||
use_tools: bool = True,
|
||||
stream: bool = False,
|
||||
**kwargs
|
||||
) -> str:
|
||||
"""对话接口。"""
|
||||
try:
|
||||
# 构建消息列表
|
||||
messages = []
|
||||
|
||||
# 系统提示词
|
||||
if system_prompt is None:
|
||||
system_prompt = self.personality.get_system_prompt()
|
||||
|
||||
# 注入记忆上下文
|
||||
if use_memory:
|
||||
short_term, long_term = await self.memory.get_context(
|
||||
user_id=user_id,
|
||||
query=user_message
|
||||
)
|
||||
|
||||
if short_term or long_term:
|
||||
memory_context = self.memory.format_context(short_term, long_term)
|
||||
system_prompt += f"\n\n{memory_context}"
|
||||
|
||||
messages.append(Message(role="system", content=system_prompt))
|
||||
|
||||
# 添加用户消息
|
||||
messages.append(Message(role="user", content=user_message))
|
||||
|
||||
# 准备工具
|
||||
tools = None
|
||||
if use_tools and self.tools.list():
|
||||
tools = self.tools.to_openai_format()
|
||||
|
||||
# 调用模型
|
||||
if stream:
|
||||
return self._chat_stream(messages, tools, **kwargs)
|
||||
else:
|
||||
response = await self.model.chat(messages, tools, **kwargs)
|
||||
|
||||
# 处理工具调用
|
||||
if response.tool_calls:
|
||||
response = await self._handle_tool_calls(
|
||||
messages, response, tools, **kwargs
|
||||
)
|
||||
|
||||
# 写入记忆
|
||||
if use_memory:
|
||||
stored_memory = await self.memory.add_qa_pair(
|
||||
user_id=user_id,
|
||||
question=user_message,
|
||||
answer=response.content,
|
||||
metadata={"source": "chat"},
|
||||
)
|
||||
if stored_memory:
|
||||
logger.info(
|
||||
"已写入长期记忆问答对:\n"
|
||||
f"{stored_memory.content}\n"
|
||||
f"memory_id={stored_memory.id}, "
|
||||
f"importance={stored_memory.importance:.2f}"
|
||||
)
|
||||
|
||||
return response.content
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"对话失败: {e}")
|
||||
raise
|
||||
|
||||
async def _chat_stream(
|
||||
self,
|
||||
messages: List[Message],
|
||||
tools: Optional[List[Dict]],
|
||||
**kwargs
|
||||
) -> AsyncIterator[str]:
|
||||
"""流式对话。"""
|
||||
async for chunk in self.model.chat_stream(messages, tools, **kwargs):
|
||||
yield chunk
|
||||
|
||||
async def _handle_tool_calls(
|
||||
self,
|
||||
messages: List[Message],
|
||||
response: Message,
|
||||
tools: Optional[List[Dict]],
|
||||
**kwargs
|
||||
) -> Message:
|
||||
"""处理工具调用。"""
|
||||
messages.append(response)
|
||||
|
||||
# 鎵ц宸ュ叿璋冪敤
|
||||
for tool_call in response.tool_calls or []:
|
||||
try:
|
||||
tool_name, tool_args, tool_call_id = self._parse_tool_call(tool_call)
|
||||
except Exception as e:
|
||||
logger.warning(f"解析工具调用失败: {e}")
|
||||
fallback_id = tool_call.get('id') if isinstance(tool_call, dict) else getattr(tool_call, 'id', None)
|
||||
if fallback_id:
|
||||
messages.append(Message(
|
||||
role="tool",
|
||||
content=f"工具参数解析失败: {str(e)}",
|
||||
tool_call_id=fallback_id
|
||||
))
|
||||
continue
|
||||
if not tool_name:
|
||||
logger.warning(f"跳过无效工具调用: {tool_call}")
|
||||
continue
|
||||
|
||||
tool_def = self.tools.get(tool_name)
|
||||
if not tool_def:
|
||||
error_msg = f"未找到工具: {tool_name}"
|
||||
logger.warning(error_msg)
|
||||
if tool_call_id:
|
||||
messages.append(Message(
|
||||
role="tool",
|
||||
content=error_msg,
|
||||
tool_call_id=tool_call_id
|
||||
))
|
||||
continue
|
||||
|
||||
try:
|
||||
result = tool_def.function(**tool_args)
|
||||
if inspect.isawaitable(result):
|
||||
result = await result
|
||||
if tool_call_id:
|
||||
messages.append(Message(
|
||||
role="tool",
|
||||
content=str(result),
|
||||
tool_call_id=tool_call_id
|
||||
))
|
||||
except Exception as e:
|
||||
if tool_call_id:
|
||||
messages.append(Message(
|
||||
role="tool",
|
||||
content=f"工具执行失败: {str(e)}",
|
||||
tool_call_id=tool_call_id
|
||||
))
|
||||
|
||||
# 再次调用模型获取最终响应
|
||||
return await self.model.chat(messages, tools, **kwargs)
|
||||
|
||||
def _parse_tool_call(self, tool_call: Any) -> Tuple[Optional[str], Dict[str, Any], Optional[str]]:
|
||||
"""兼容不同 SDK 返回的工具调用结构。"""
|
||||
if isinstance(tool_call, dict):
|
||||
tool_call_id = tool_call.get('id')
|
||||
function = tool_call.get('function') or {}
|
||||
tool_name = function.get('name')
|
||||
raw_args = function.get('arguments')
|
||||
else:
|
||||
tool_call_id = getattr(tool_call, 'id', None)
|
||||
function = getattr(tool_call, 'function', None)
|
||||
tool_name = getattr(function, 'name', None) if function else None
|
||||
raw_args = getattr(function, 'arguments', None) if function else None
|
||||
|
||||
tool_args = self._normalize_tool_args(raw_args)
|
||||
return tool_name, tool_args, tool_call_id
|
||||
|
||||
def _normalize_tool_args(self, raw_args: Any) -> Dict[str, Any]:
|
||||
"""将工具参数统一转换为字典。"""
|
||||
if raw_args is None:
|
||||
return {}
|
||||
|
||||
if isinstance(raw_args, dict):
|
||||
return raw_args
|
||||
|
||||
if isinstance(raw_args, str):
|
||||
raw_args = raw_args.strip()
|
||||
if not raw_args:
|
||||
return {}
|
||||
parsed = json.loads(raw_args)
|
||||
if not isinstance(parsed, dict):
|
||||
raise ValueError(f"工具参数必须是 JSON 对象，实际类型: {type(parsed)}")
|
||||
return parsed
|
||||
|
||||
if hasattr(raw_args, 'model_dump'):
|
||||
parsed = raw_args.model_dump()
|
||||
if isinstance(parsed, dict):
|
||||
return parsed
|
||||
|
||||
raise ValueError(f"不支持的工具参数类型: {type(raw_args)}")
|
||||
|
||||
def set_personality(self, personality_name: str) -> bool:
|
||||
"""设置人格。"""
|
||||
return self.personality.set_personality(personality_name)
|
||||
|
||||
def list_personalities(self) -> List[str]:
|
||||
"""列出所有人格。"""
|
||||
return self.personality.list_personalities()
|
||||
|
||||
def switch_model(self, model_config: ModelConfig) -> bool:
|
||||
"""Runtime switch for primary chat model."""
|
||||
new_model = self._create_model(model_config)
|
||||
self.model = new_model
|
||||
self.config = model_config
|
||||
logger.info(
|
||||
f"已切换主模型: {model_config.provider.value}/{model_config.model_name}"
|
||||
)
|
||||
return True
|
||||
|
||||
async def create_long_task(
|
||||
self,
|
||||
user_id: str,
|
||||
title: str,
|
||||
description: str,
|
||||
steps: List[Dict],
|
||||
metadata: Optional[Dict] = None
|
||||
) -> str:
|
||||
"""创建长任务。"""
|
||||
return self.task_manager.create_task(
|
||||
user_id=user_id,
|
||||
title=title,
|
||||
description=description,
|
||||
steps=steps,
|
||||
metadata=metadata
|
||||
)
|
||||
|
||||
async def start_task(
|
||||
self,
|
||||
task_id: str,
|
||||
progress_callback: Optional[callable] = None
|
||||
):
|
||||
"""启动任务。"""
|
||||
await self.task_manager.start_task(task_id, progress_callback)
|
||||
|
||||
def get_task_status(self, task_id: str) -> Optional[Dict]:
|
||||
"""获取任务状态。"""
|
||||
return self.task_manager.get_task_status(task_id)
|
||||
|
||||
def register_tool(self, name: str, description: str, parameters: Dict, function: callable):
|
||||
"""注册工具。"""
|
||||
from .base import ToolDefinition
|
||||
tool = ToolDefinition(
|
||||
name=name,
|
||||
description=description,
|
||||
parameters=parameters,
|
||||
function=function
|
||||
)
|
||||
self.tools.register(tool)
|
||||
logger.info(f"已注册工具: {name}")
|
||||
|
||||
def unregister_tool(self, name: str) -> bool:
|
||||
"""卸载工具。"""
|
||||
removed = self.tools.unregister(name)
|
||||
if removed:
|
||||
logger.info(f"已卸载工具: {name}")
|
||||
return removed
|
||||
|
||||
def unregister_tools_by_prefix(self, prefix: str) -> int:
|
||||
"""按前缀批量卸载工具。"""
|
||||
removed_count = self.tools.unregister_by_prefix(prefix)
|
||||
if removed_count:
|
||||
logger.info(f"Unregistered tools by prefix {prefix}: {removed_count}")
|
||||
return removed_count
|
||||
|
||||
def clear_memory(self, user_id: str):
|
||||
"""清除用户短期记忆。"""
|
||||
self.memory.clear_short_term(user_id)
|
||||
logger.info(f"Cleared short-term memory for user {user_id}")
|
||||
|
||||
async def clear_long_term_memory(self, user_id: str) -> bool:
|
||||
try:
|
||||
await self.memory.clear_long_term(user_id)
|
||||
logger.info(f"Cleared long-term memory for user {user_id}")
|
||||
return True
|
||||
except Exception as e:
|
||||
logger.warning(f"Failed to clear long-term memory for user {user_id}: {e}")
|
||||
return False
|
||||
|
||||
async def list_long_term_memories(self, user_id: str, limit: int = 20):
|
||||
return await self.memory.list_long_term(user_id, limit=limit)
|
||||
|
||||
async def get_long_term_memory(self, user_id: str, memory_id: str):
|
||||
return await self.memory.get_long_term(user_id, memory_id)
|
||||
|
||||
async def add_long_term_memory(
|
||||
self,
|
||||
user_id: str,
|
||||
content: str,
|
||||
importance: float = 0.8,
|
||||
metadata: Optional[Dict] = None,
|
||||
):
|
||||
return await self.memory.add_long_term(
|
||||
user_id=user_id,
|
||||
content=content,
|
||||
importance=importance,
|
||||
metadata=metadata,
|
||||
)
|
||||
|
||||
async def search_long_term_memories(
|
||||
self, user_id: str, query: str, limit: int = 10
|
||||
):
|
||||
return await self.memory.search_long_term(user_id, query=query, limit=limit)
|
||||
|
||||
async def update_long_term_memory(
|
||||
self,
|
||||
user_id: str,
|
||||
memory_id: str,
|
||||
content: Optional[str] = None,
|
||||
importance: Optional[float] = None,
|
||||
metadata: Optional[Dict] = None,
|
||||
):
|
||||
return await self.memory.update_long_term(
|
||||
user_id=user_id,
|
||||
memory_id=memory_id,
|
||||
content=content,
|
||||
importance=importance,
|
||||
metadata=metadata,
|
||||
)
|
||||
|
||||
async def delete_long_term_memory(self, user_id: str, memory_id: str) -> bool:
|
||||
return await self.memory.delete_long_term(user_id, memory_id)
|
||||
|
||||
async def clear_all_memory(self, user_id: str) -> bool:
|
||||
"""清除用户全部记忆（短期 + 长期）。"""
|
||||
self.clear_memory(user_id)
|
||||
try:
|
||||
return await self.clear_long_term_memory(user_id)
|
||||
except Exception:
|
||||
return False
|
||||
|
||||
545
src/ai/docs/README.md
Normal file
545
src/ai/docs/README.md
Normal file
@@ -0,0 +1,545 @@
|
||||
# AI模型接入完整指南
|
||||
|
||||
## 概述
|
||||
|
||||
本系统提供了完整的AI模型接入能力，包括：
|
||||
|
||||
- ✅ **多模型支持**：OpenAI、Anthropic Claude、DeepSeek、Qwen等
|
||||
- ✅ **人格系统**：可自定义AI性格特征和说话风格
|
||||
- ✅ **记忆系统**：短期记忆（滑动窗口）+ 长期记忆（RAG检索）
|
||||
- ✅ **长任务处理**：支持多步骤复杂任务的执行和管理
|
||||
- ✅ **Skills插件**：可扩展的技能插件系统
|
||||
- ✅ **MCP支持**：Model Context Protocol标准支持
|
||||
|
||||
## 快速开始
|
||||
|
||||
### 1. 安装依赖
|
||||
|
||||
```bash
|
||||
pip install openai anthropic numpy
|
||||
```
|
||||
|
||||
### 2. 配置环境变量
|
||||
|
||||
在 `.env` 文件中添加：
|
||||
|
||||
```env
|
||||
# AI模型配置
|
||||
AI_PROVIDER=openai # openai, anthropic, deepseek, qwen
|
||||
AI_MODEL=gpt-4
|
||||
AI_API_KEY=your_api_key_here
|
||||
AI_API_BASE=https://api.openai.com/v1 # 可选，自定义API地址
|
||||
```
|
||||
|
||||
### 3. 基础使用
|
||||
|
||||
```python
|
||||
from pathlib import Path
|
||||
from src.ai import AIClient
|
||||
from src.ai.base import ModelConfig, ModelProvider
|
||||
|
||||
# 创建模型配置
|
||||
config = ModelConfig(
|
||||
provider=ModelProvider.OPENAI,
|
||||
model_name="gpt-4",
|
||||
api_key="your_api_key",
|
||||
temperature=0.7
|
||||
)
|
||||
|
||||
# 初始化AI客户端
|
||||
client = AIClient(config, data_dir=Path("data/ai"))
|
||||
|
||||
# 对话
|
||||
response = await client.chat(
|
||||
user_id="user123",
|
||||
user_message="你好，介绍一下你自己",
|
||||
use_memory=True,
|
||||
use_tools=True
|
||||
)
|
||||
|
||||
print(response)
|
||||
```
|
||||
|
||||
## 核心功能详解
|
||||
|
||||
### 1. 人格系统
|
||||
|
||||
人格系统允许你为AI定义独特的性格特征和说话风格。
|
||||
|
||||
#### 使用预设人格
|
||||
|
||||
```python
|
||||
# 列出所有人格
|
||||
personalities = client.list_personalities()
|
||||
print(personalities) # ['default', 'tech_expert', 'creative']
|
||||
|
||||
# 切换人格
|
||||
client.set_personality('tech_expert')
|
||||
```
|
||||
|
||||
#### 自定义人格
|
||||
|
||||
```python
|
||||
from src.ai.personality import PersonalityProfile, PersonalityTrait
|
||||
|
||||
# 创建自定义人格
|
||||
custom_personality = PersonalityProfile(
|
||||
name="游戏助手",
|
||||
description="专业的游戏顾问和陪玩伙伴",
|
||||
traits=[
|
||||
PersonalityTrait.FRIENDLY,
|
||||
PersonalityTrait.HUMOROUS,
|
||||
PersonalityTrait.CREATIVE
|
||||
],
|
||||
speaking_style="活泼、热情、善用游戏术语和梗",
|
||||
example_responses=[
|
||||
"哇！这个操作太秀了！",
|
||||
"让我们一起冲冲冲！",
|
||||
"别慌，稳住我们能赢！"
|
||||
],
|
||||
custom_instructions="始终保持积极乐观的态度，鼓励玩家"
|
||||
)
|
||||
|
||||
# 添加到系统
|
||||
client.personality.add_personality('gamer', custom_personality)
|
||||
client.set_personality('gamer')
|
||||
```
|
||||
|
||||
### 2. 记忆系统
|
||||
|
||||
记忆系统分为短期记忆和长期记忆两部分。
|
||||
|
||||
#### 短期记忆
|
||||
|
||||
- 自动保存最近的对话（默认20条）
|
||||
- 使用滑动窗口，自动清理过期记忆（默认30分钟）
|
||||
- 无需手动管理
|
||||
|
||||
#### 长期记忆
|
||||
|
||||
- 自动评估重要性，重要对话会被保存
|
||||
- 支持向量检索（RAG）
|
||||
- 可手动查询历史记忆
|
||||
|
||||
```python
|
||||
# 获取用户的记忆上下文
|
||||
short_term, long_term = await client.memory.get_context(
|
||||
user_id="user123",
|
||||
query="我们之前讨论过什么？",
|
||||
max_short_term=10,
|
||||
max_long_term=5
|
||||
)
|
||||
|
||||
# 格式化为文本
|
||||
context = client.memory.format_context(short_term, long_term)
|
||||
print(context)
|
||||
|
||||
# 清除短期记忆
|
||||
client.clear_memory("user123")
|
||||
```
|
||||
|
||||
#### 记忆重要性评估
|
||||
|
||||
系统会自动评估对话的重要性：
|
||||
|
||||
- 内容长度（>100字符 +0.1分）
|
||||
- 关键词（"重要"、"记住"、"别忘了" +0.2分）
|
||||
- 元数据（问题 +0.1分，代码 +0.15分）
|
||||
- 阈值：≥0.6分会被保存到长期记忆
|
||||
|
||||
### 3. 长任务处理
|
||||
|
||||
支持将复杂任务分解为多个步骤执行。
|
||||
|
||||
```python
|
||||
# 创建长任务
|
||||
task_id = await client.create_long_task(
|
||||
user_id="user123",
|
||||
title="数据分析任务",
|
||||
description="分析用户行为数据并生成报告",
|
||||
steps=[
|
||||
{
|
||||
"description": "加载数据",
|
||||
"action": "load_data",
|
||||
"params": {"file_path": "data.csv"}
|
||||
},
|
||||
{
|
||||
"description": "数据清洗",
|
||||
"action": "clean_data",
|
||||
"params": {}
|
||||
},
|
||||
{
|
||||
"description": "统计分析",
|
||||
"action": "analyze",
|
||||
"params": {"metrics": ["avg", "max", "min"]}
|
||||
},
|
||||
{
|
||||
"description": "生成报告",
|
||||
"action": "generate_report",
|
||||
"params": {"format": "pdf"}
|
||||
}
|
||||
]
|
||||
)
|
||||
|
||||
# 定义进度回调
|
||||
async def on_progress(task_id, progress, message):
|
||||
print(f"任务 {task_id}: {progress*100:.1f}% - {message}")
|
||||
|
||||
# 启动任务
|
||||
await client.start_task(task_id, progress_callback=on_progress)
|
||||
|
||||
# 查询任务状态
|
||||
status = client.get_task_status(task_id)
|
||||
print(status)
|
||||
```
|
||||
|
||||
#### 注册任务动作处理器
|
||||
|
||||
```python
|
||||
# 注册动作处理器
|
||||
async def load_data_handler(file_path: str):
|
||||
# 实现数据加载逻辑
|
||||
return {"rows": 1000, "columns": 10}
|
||||
|
||||
client.task_manager.register_action("load_data", load_data_handler)
|
||||
```
|
||||
|
||||
### 4. Skills插件系统
|
||||
|
||||
Skills系统允许你扩展AI的能力。
|
||||
|
||||
#### 加载技能
|
||||
|
||||
```python
|
||||
from pathlib import Path
|
||||
from src.ai.skills import SkillsManager
|
||||
|
||||
# 创建技能管理器
|
||||
skills_manager = SkillsManager(skills_dir=Path("skills"))
|
||||
|
||||
# 加载单个技能
|
||||
await skills_manager.load_skill("weather")
|
||||
|
||||
# 加载所有技能
|
||||
await skills_manager.load_all_skills()
|
||||
|
||||
# 获取所有工具
|
||||
tools = skills_manager.get_all_tools()
|
||||
```
|
||||
|
||||
#### 创建自定义技能
|
||||
|
||||
```python
|
||||
from src.ai.skills import create_skill_template
|
||||
|
||||
# 创建技能模板
|
||||
create_skill_template("my_skill", Path("skills"))
|
||||
```
|
||||
|
||||
技能目录结构：
|
||||
|
||||
```
|
||||
skills/
|
||||
└── my_skill/
|
||||
├── skill.json # 元数据
|
||||
├── main.py # 主要实现
|
||||
├── __init__.py
|
||||
└── README.md
|
||||
```
|
||||
|
||||
`main.py` 示例：
|
||||
|
||||
```python
|
||||
from src.ai.skills.base import Skill
|
||||
|
||||
class MySkill(Skill):
|
||||
async def initialize(self):
|
||||
# 注册工具
|
||||
self.register_tool("my_tool", self.my_tool)
|
||||
|
||||
async def my_tool(self, param: str) -> str:
|
||||
return f"处理: {param}"
|
||||
|
||||
async def cleanup(self):
|
||||
pass
|
||||
```
|
||||
|
||||
### 5. MCP支持
|
||||
|
||||
MCP (Model Context Protocol) 提供标准化的上下文访问接口。
|
||||
|
||||
#### 使用内置MCP服务器
|
||||
|
||||
```python
|
||||
from pathlib import Path
|
||||
from src.ai.mcp import MCPManager
|
||||
from src.ai.mcp.servers import FileSystemMCPServer
|
||||
|
||||
# 创建MCP管理器
|
||||
mcp_manager = MCPManager(config_path=Path("config/mcp.json"))
|
||||
|
||||
# 注册文件系统服务器
|
||||
fs_server = FileSystemMCPServer(root_path=Path("data"))
|
||||
await mcp_manager.register_server(fs_server)
|
||||
|
||||
# 获取所有工具
|
||||
tools = await mcp_manager.get_all_tools_for_ai()
|
||||
|
||||
# 执行工具
|
||||
result = await mcp_manager.execute_tool(
|
||||
"filesystem.read_file",
|
||||
{"path": "test.txt"}
|
||||
)
|
||||
```
|
||||
|
||||
#### 创建自定义MCP服务器
|
||||
|
||||
```python
|
||||
from src.ai.mcp.base import MCPServer
|
||||
|
||||
class MyMCPServer(MCPServer):
|
||||
def __init__(self):
|
||||
super().__init__(name="my_server", version="1.0.0")
|
||||
|
||||
async def initialize(self):
|
||||
self.register_tool(
|
||||
name="my_tool",
|
||||
description="我的工具",
|
||||
input_schema={
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"param": {"type": "string"}
|
||||
},
|
||||
"required": ["param"]
|
||||
},
|
||||
handler=self.my_tool_handler
|
||||
)
|
||||
|
||||
async def my_tool_handler(self, param: str):
|
||||
return f"处理: {param}"
|
||||
```
|
||||
|
||||
## 集成到QQ机器人
|
||||
|
||||
### 修改消息处理器
|
||||
|
||||
```python
|
||||
# src/handlers/message_handler.py
|
||||
from src.ai import AIClient
|
||||
from src.ai.base import ModelConfig, ModelProvider
|
||||
from pathlib import Path
|
||||
|
||||
class MessageHandler:
|
||||
def __init__(self, bot):
|
||||
self.bot = bot
|
||||
|
||||
# 初始化AI客户端
|
||||
config = ModelConfig(
|
||||
provider=ModelProvider.OPENAI,
|
||||
model_name="gpt-4",
|
||||
api_key=Config.AI_API_KEY,
|
||||
api_base=Config.AI_API_BASE
|
||||
)
|
||||
self.ai_client = AIClient(config, data_dir=Path("data/ai"))
|
||||
|
||||
# 加载技能
|
||||
from src.ai.skills import SkillsManager
|
||||
self.skills = SkillsManager(Path("skills"))
|
||||
asyncio.create_task(self.skills.load_all_skills())
|
||||
|
||||
async def handle_at_message(self, message: Message):
|
||||
try:
|
||||
# 提取用户消息
|
||||
user_message = message.content.strip()
|
||||
user_id = message.author.id
|
||||
|
||||
# 调用AI
|
||||
response = await self.ai_client.chat(
|
||||
user_id=user_id,
|
||||
user_message=user_message,
|
||||
use_memory=True,
|
||||
use_tools=True
|
||||
)
|
||||
|
||||
# 发送回复
|
||||
await message.reply(content=response)
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"处理消息失败: {e}")
|
||||
await message.reply(content="抱歉，处理消息时出错了")
|
||||
```
|
||||
|
||||
## 配置文件
|
||||
|
||||
### config/personalities.json
|
||||
|
||||
```json
|
||||
{
|
||||
"default": {
|
||||
"name": "小助手",
|
||||
"description": "友好的AI助手",
|
||||
"traits": ["FRIENDLY", "PROFESSIONAL", "EMPATHETIC"],
|
||||
"speaking_style": "温和、清晰、有条理",
|
||||
"example_responses": [
|
||||
"我明白你的意思了，让我来帮你解决这个问题。",
|
||||
"这是个很好的问题！我来详细解释一下。"
|
||||
],
|
||||
"custom_instructions": ""
|
||||
}
|
||||
}
|
||||
```
|
||||
|
||||
### config/mcp.json
|
||||
|
||||
```json
|
||||
{
|
||||
"filesystem": {
|
||||
"enabled": true,
|
||||
"root_path": "data"
|
||||
}
|
||||
}
|
||||
```
|
||||
|
||||
## 高级用法
|
||||
|
||||
### 1. 流式响应
|
||||
|
||||
```python
|
||||
async for chunk in client.chat(
|
||||
user_id="user123",
|
||||
user_message="讲个故事",
|
||||
stream=True
|
||||
):
|
||||
print(chunk, end='', flush=True)
|
||||
```
|
||||
|
||||
### 2. 自定义工具
|
||||
|
||||
```python
|
||||
# 定义工具函数
|
||||
async def search_web(query: str) -> str:
|
||||
# 实现搜索逻辑
|
||||
return f"搜索结果: {query}"
|
||||
|
||||
# 注册工具
|
||||
client.register_tool(
|
||||
name="search_web",
|
||||
description="搜索网络信息",
|
||||
parameters={
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"query": {
|
||||
"type": "string",
|
||||
"description": "搜索关键词"
|
||||
}
|
||||
},
|
||||
"required": ["query"]
|
||||
},
|
||||
function=search_web
|
||||
)
|
||||
```
|
||||
|
||||
### 3. 多模型切换
|
||||
|
||||
```python
|
||||
# 切换到Claude
|
||||
claude_config = ModelConfig(
|
||||
provider=ModelProvider.ANTHROPIC,
|
||||
model_name="claude-3-5-sonnet-20241022",
|
||||
api_key="your_anthropic_key"
|
||||
)
|
||||
client.model = client._create_model(claude_config)
|
||||
|
||||
# 切换到DeepSeek
|
||||
deepseek_config = ModelConfig(
|
||||
provider=ModelProvider.DEEPSEEK,
|
||||
model_name="deepseek-chat",
|
||||
api_key="your_deepseek_key",
|
||||
api_base="https://api.deepseek.com/v1"
|
||||
)
|
||||
client.model = client._create_model(deepseek_config)
|
||||
```
|
||||
|
||||
## 性能优化
|
||||
|
||||
### 1. 记忆系统优化
|
||||
|
||||
```python
|
||||
# 调整短期记忆大小
|
||||
client.memory.short_term.max_size = 30 # 默认20
|
||||
|
||||
# 调整记忆过期时间
|
||||
from datetime import timedelta
|
||||
client.memory.short_term.max_age = timedelta(hours=1) # 默认30分钟
|
||||
|
||||
# 调整重要性阈值
|
||||
client.memory.importance_threshold = 0.7 # 默认0.6
|
||||
```
|
||||
|
||||
### 2. 并发任务处理
|
||||
|
||||
```python
|
||||
# 同时启动多个任务
|
||||
tasks = []
|
||||
for i in range(5):
|
||||
task_id = await client.create_long_task(...)
|
||||
tasks.append(client.start_task(task_id))
|
||||
|
||||
# 等待所有任务完成
|
||||
await asyncio.gather(*tasks)
|
||||
```
|
||||
|
||||
## 故障排查
|
||||
|
||||
### 常见问题
|
||||
|
||||
1. **API调用失败**
|
||||
- 检查API密钥是否正确
|
||||
- 检查网络连接
|
||||
- 检查API地址是否正确
|
||||
|
||||
2. **记忆系统不工作**
|
||||
- 确保 `use_memory=True`
|
||||
- 检查数据目录权限
|
||||
- 查看日志文件
|
||||
|
||||
3. **技能加载失败**
|
||||
- 检查 `skill.json` 格式
|
||||
- 确保 `main.py` 中有Skill子类
|
||||
- 查看错误日志
|
||||
|
||||
## 最佳实践
|
||||
|
||||
1. **合理使用记忆系统**
|
||||
- 不要在每次对话都查询长期记忆
|
||||
- 定期清理不重要的记忆
|
||||
- 为重要对话添加元数据标记
|
||||
|
||||
2. **技能开发**
|
||||
- 保持技能功能单一
|
||||
- 提供清晰的工具描述
|
||||
- 处理异常情况
|
||||
|
||||
3. **长任务设计**
|
||||
- 将任务分解为合理的步骤
|
||||
- 提供进度反馈
|
||||
- 支持任务暂停和恢复
|
||||
|
||||
## 更新日志
|
||||
|
||||
### v1.0.0 (2024-03-02)
|
||||
- ✅ 初始版本发布
|
||||
- ✅ 支持OpenAI、Anthropic、DeepSeek、Qwen
|
||||
- ✅ 人格系统
|
||||
- ✅ 记忆系统（短期+长期+RAG）
|
||||
- ✅ 长任务处理
|
||||
- ✅ Skills插件系统
|
||||
- ✅ MCP支持
|
||||
|
||||
## 贡献
|
||||
|
||||
欢迎提交Issue和Pull Request！
|
||||
|
||||
## 许可证
|
||||
|
||||
MIT License
|
||||
13
src/ai/mcp/__init__.py
Normal file
13
src/ai/mcp/__init__.py
Normal file
@@ -0,0 +1,13 @@
|
||||
"""
|
||||
MCP模块
|
||||
"""
|
||||
from .base import MCPServer, MCPClient, MCPManager, MCPResource, MCPTool, MCPPrompt
|
||||
|
||||
__all__ = [
|
||||
'MCPServer',
|
||||
'MCPClient',
|
||||
'MCPManager',
|
||||
'MCPResource',
|
||||
'MCPTool',
|
||||
'MCPPrompt'
|
||||
]
|
||||
219
src/ai/mcp/base.py
Normal file
219
src/ai/mcp/base.py
Normal file
@@ -0,0 +1,219 @@
|
||||
"""
|
||||
MCP (Model Context Protocol) 支持
|
||||
"""
|
||||
import asyncio
|
||||
import json
|
||||
from typing import Dict, List, Optional, Any, Callable
|
||||
from dataclasses import dataclass, asdict
|
||||
from pathlib import Path
|
||||
from src.utils.logger import setup_logger
|
||||
|
||||
logger = setup_logger('MCPSystem')
|
||||
|
||||
|
||||
@dataclass
|
||||
class MCPResource:
|
||||
"""MCP资源"""
|
||||
uri: str
|
||||
name: str
|
||||
description: str
|
||||
mime_type: str
|
||||
|
||||
|
||||
@dataclass
|
||||
class MCPTool:
|
||||
"""MCP工具"""
|
||||
name: str
|
||||
description: str
|
||||
input_schema: Dict[str, Any]
|
||||
|
||||
|
||||
@dataclass
|
||||
class MCPPrompt:
|
||||
"""MCP提示词"""
|
||||
name: str
|
||||
description: str
|
||||
arguments: List[Dict[str, Any]]
|
||||
|
||||
|
||||
class MCPServer:
|
||||
"""MCP服务器基类"""
|
||||
|
||||
def __init__(self, name: str, version: str):
|
||||
self.name = name
|
||||
self.version = version
|
||||
self.resources: Dict[str, MCPResource] = {}
|
||||
self.tools: Dict[str, Callable] = {}
|
||||
self.prompts: Dict[str, MCPPrompt] = {}
|
||||
|
||||
async def initialize(self):
|
||||
"""初始化服务器"""
|
||||
pass
|
||||
|
||||
async def shutdown(self):
|
||||
"""关闭服务器"""
|
||||
pass
|
||||
|
||||
def register_resource(self, resource: MCPResource):
|
||||
"""注册资源"""
|
||||
self.resources[resource.uri] = resource
|
||||
|
||||
def register_tool(self, name: str, description: str, input_schema: Dict, handler: Callable):
|
||||
"""注册工具"""
|
||||
tool = MCPTool(name=name, description=description, input_schema=input_schema)
|
||||
self.tools[name] = handler
|
||||
logger.info(f"✅ MCP工具注册: {self.name}.{name}")
|
||||
|
||||
def register_prompt(self, prompt: MCPPrompt):
|
||||
"""注册提示词"""
|
||||
self.prompts[prompt.name] = prompt
|
||||
|
||||
async def list_resources(self) -> List[MCPResource]:
|
||||
"""列出资源"""
|
||||
return list(self.resources.values())
|
||||
|
||||
async def read_resource(self, uri: str) -> Optional[str]:
|
||||
"""读取资源"""
|
||||
raise NotImplementedError
|
||||
|
||||
async def list_tools(self) -> List[MCPTool]:
|
||||
"""列出工具"""
|
||||
return [
|
||||
MCPTool(name=name, description="", input_schema={})
|
||||
for name in self.tools.keys()
|
||||
]
|
||||
|
||||
async def call_tool(self, name: str, arguments: Dict[str, Any]) -> Any:
|
||||
"""调用工具"""
|
||||
if name not in self.tools:
|
||||
raise ValueError(f"工具不存在: {name}")
|
||||
|
||||
handler = self.tools[name]
|
||||
return await handler(**arguments)
|
||||
|
||||
async def list_prompts(self) -> List[MCPPrompt]:
|
||||
"""列出提示词"""
|
||||
return list(self.prompts.values())
|
||||
|
||||
async def get_prompt(self, name: str, arguments: Dict[str, Any]) -> Optional[str]:
|
||||
"""获取提示词"""
|
||||
raise NotImplementedError
|
||||
|
||||
|
||||
class MCPClient:
|
||||
"""MCP客户端"""
|
||||
|
||||
def __init__(self):
|
||||
self.servers: Dict[str, MCPServer] = {}
|
||||
|
||||
async def connect_server(self, server: MCPServer):
|
||||
"""连接服务器"""
|
||||
await server.initialize()
|
||||
self.servers[server.name] = server
|
||||
logger.info(f"✅ 连接MCP服务器: {server.name} v{server.version}")
|
||||
|
||||
async def disconnect_server(self, server_name: str):
|
||||
"""断开服务器"""
|
||||
if server_name in self.servers:
|
||||
await self.servers[server_name].shutdown()
|
||||
del self.servers[server_name]
|
||||
logger.info(f"✅ 断开MCP服务器: {server_name}")
|
||||
|
||||
def get_server(self, name: str) -> Optional[MCPServer]:
|
||||
"""获取服务器"""
|
||||
return self.servers.get(name)
|
||||
|
||||
def list_servers(self) -> List[str]:
|
||||
"""列出所有服务器"""
|
||||
return list(self.servers.keys())
|
||||
|
||||
async def list_all_resources(self) -> Dict[str, List[MCPResource]]:
|
||||
"""列出所有资源"""
|
||||
result = {}
|
||||
for name, server in self.servers.items():
|
||||
result[name] = await server.list_resources()
|
||||
return result
|
||||
|
||||
async def list_all_tools(self) -> Dict[str, List[MCPTool]]:
|
||||
"""列出所有工具"""
|
||||
result = {}
|
||||
for name, server in self.servers.items():
|
||||
result[name] = await server.list_tools()
|
||||
return result
|
||||
|
||||
async def call_tool(self, server_name: str, tool_name: str, arguments: Dict[str, Any]) -> Any:
|
||||
"""调用工具"""
|
||||
server = self.get_server(server_name)
|
||||
if not server:
|
||||
raise ValueError(f"服务器不存在: {server_name}")
|
||||
|
||||
return await server.call_tool(tool_name, arguments)
|
||||
|
||||
|
||||
class MCPManager:
|
||||
"""MCP管理器"""
|
||||
|
||||
def __init__(self, config_path: Path):
|
||||
self.config_path = config_path
|
||||
self.client = MCPClient()
|
||||
self.server_configs: Dict[str, Dict] = {}
|
||||
self._load_config()
|
||||
|
||||
def _load_config(self):
|
||||
"""加载配置"""
|
||||
if self.config_path.exists():
|
||||
with open(self.config_path, 'r', encoding='utf-8') as f:
|
||||
self.server_configs = json.load(f)
|
||||
|
||||
def _save_config(self):
|
||||
"""保存配置"""
|
||||
self.config_path.parent.mkdir(parents=True, exist_ok=True)
|
||||
with open(self.config_path, 'w', encoding='utf-8') as f:
|
||||
json.dump(self.server_configs, f, ensure_ascii=False, indent=2)
|
||||
|
||||
async def register_server(self, server: MCPServer, config: Optional[Dict] = None):
|
||||
"""注册服务器"""
|
||||
await self.client.connect_server(server)
|
||||
|
||||
if config:
|
||||
self.server_configs[server.name] = config
|
||||
self._save_config()
|
||||
|
||||
async def unregister_server(self, server_name: str):
|
||||
"""注销服务器"""
|
||||
await self.client.disconnect_server(server_name)
|
||||
|
||||
if server_name in self.server_configs:
|
||||
del self.server_configs[server_name]
|
||||
self._save_config()
|
||||
|
||||
def get_client(self) -> MCPClient:
|
||||
"""获取客户端"""
|
||||
return self.client
|
||||
|
||||
async def get_all_tools_for_ai(self) -> List[Dict]:
|
||||
"""获取所有工具（AI格式）"""
|
||||
all_tools = []
|
||||
tools_by_server = await self.client.list_all_tools()
|
||||
|
||||
for server_name, tools in tools_by_server.items():
|
||||
for tool in tools:
|
||||
all_tools.append({
|
||||
"type": "function",
|
||||
"function": {
|
||||
"name": f"{server_name}.{tool.name}",
|
||||
"description": tool.description,
|
||||
"parameters": tool.input_schema
|
||||
}
|
||||
})
|
||||
|
||||
return all_tools
|
||||
|
||||
async def execute_tool(self, full_tool_name: str, arguments: Dict) -> Any:
|
||||
"""执行工具"""
|
||||
parts = full_tool_name.split('.', 1)
|
||||
if len(parts) != 2:
|
||||
raise ValueError(f"工具名格式错误: {full_tool_name}")
|
||||
|
||||
server_name, tool_name = parts
|
||||
return await self.client.call_tool(server_name, tool_name, arguments)
|
||||
6
src/ai/mcp/servers/__init__.py
Normal file
6
src/ai/mcp/servers/__init__.py
Normal file
@@ -0,0 +1,6 @@
|
||||
"""
|
||||
MCP服务器实现
|
||||
"""
|
||||
from .filesystem import FileSystemMCPServer
|
||||
|
||||
__all__ = ['FileSystemMCPServer']
|
||||
123
src/ai/mcp/servers/filesystem.py
Normal file
123
src/ai/mcp/servers/filesystem.py
Normal file
@@ -0,0 +1,123 @@
|
||||
"""
|
||||
MCP示例服务器 - 文件系统访问
|
||||
"""
|
||||
from pathlib import Path
|
||||
from typing import Optional
|
||||
from ..base import MCPServer, MCPResource
|
||||
|
||||
|
||||
class FileSystemMCPServer(MCPServer):
|
||||
"""文件系统MCP服务器"""
|
||||
|
||||
def __init__(self, root_path: Path):
|
||||
super().__init__(name="filesystem", version="1.0.0")
|
||||
self.root_path = root_path
|
||||
|
||||
async def initialize(self):
|
||||
"""初始化"""
|
||||
# 注册工具
|
||||
self.register_tool(
|
||||
name="read_file",
|
||||
description="读取文件内容",
|
||||
input_schema={
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"path": {
|
||||
"type": "string",
|
||||
"description": "文件路径"
|
||||
}
|
||||
},
|
||||
"required": ["path"]
|
||||
},
|
||||
handler=self.read_file
|
||||
)
|
||||
|
||||
self.register_tool(
|
||||
name="write_file",
|
||||
description="写入文件内容",
|
||||
input_schema={
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"path": {
|
||||
"type": "string",
|
||||
"description": "文件路径"
|
||||
},
|
||||
"content": {
|
||||
"type": "string",
|
||||
"description": "文件内容"
|
||||
}
|
||||
},
|
||||
"required": ["path", "content"]
|
||||
},
|
||||
handler=self.write_file
|
||||
)
|
||||
|
||||
self.register_tool(
|
||||
name="list_directory",
|
||||
description="列出目录内容",
|
||||
input_schema={
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"path": {
|
||||
"type": "string",
|
||||
"description": "目录路径"
|
||||
}
|
||||
},
|
||||
"required": ["path"]
|
||||
},
|
||||
handler=self.list_directory
|
||||
)
|
||||
|
||||
def _resolve_path(self, path: str) -> Path:
|
||||
"""解析路径"""
|
||||
full_path = (self.root_path / path).resolve()
|
||||
|
||||
# 安全检查：确保路径在root_path内
|
||||
if not str(full_path).startswith(str(self.root_path)):
|
||||
raise ValueError("路径超出允许范围")
|
||||
|
||||
return full_path
|
||||
|
||||
async def read_file(self, path: str) -> str:
|
||||
"""读取文件"""
|
||||
file_path = self._resolve_path(path)
|
||||
|
||||
if not file_path.exists():
|
||||
raise FileNotFoundError(f"文件不存在: {path}")
|
||||
|
||||
if not file_path.is_file():
|
||||
raise ValueError(f"不是文件: {path}")
|
||||
|
||||
with open(file_path, 'r', encoding='utf-8') as f:
|
||||
return f.read()
|
||||
|
||||
async def write_file(self, path: str, content: str) -> str:
|
||||
"""写入文件"""
|
||||
file_path = self._resolve_path(path)
|
||||
|
||||
file_path.parent.mkdir(parents=True, exist_ok=True)
|
||||
|
||||
with open(file_path, 'w', encoding='utf-8') as f:
|
||||
f.write(content)
|
||||
|
||||
return f"文件已写入: {path}"
|
||||
|
||||
async def list_directory(self, path: str) -> list:
|
||||
"""列出目录"""
|
||||
dir_path = self._resolve_path(path)
|
||||
|
||||
if not dir_path.exists():
|
||||
raise FileNotFoundError(f"目录不存在: {path}")
|
||||
|
||||
if not dir_path.is_dir():
|
||||
raise ValueError(f"不是目录: {path}")
|
||||
|
||||
items = []
|
||||
for item in dir_path.iterdir():
|
||||
items.append({
|
||||
"name": item.name,
|
||||
"type": "directory" if item.is_dir() else "file",
|
||||
"size": item.stat().st_size if item.is_file() else None
|
||||
})
|
||||
|
||||
return items
|
||||
466
src/ai/memory.py
Normal file
466
src/ai/memory.py
Normal file
@@ -0,0 +1,466 @@
|
||||
"""
|
||||
记忆系统：短期记忆、长期记忆与 RAG 检索（向量数据库）。
|
||||
"""
|
||||
import asyncio
|
||||
import hashlib
|
||||
import uuid
|
||||
from typing import List, Dict, Optional, Tuple, Callable, Awaitable
|
||||
from dataclasses import dataclass, field
|
||||
from datetime import datetime, timedelta
|
||||
from pathlib import Path
|
||||
from collections import deque
|
||||
from .vector_store import VectorStore, VectorMemory, ChromaVectorStore, JSONVectorStore
|
||||
from src.utils.logger import setup_logger
|
||||
|
||||
logger = setup_logger('MemorySystem')
|
||||
|
||||
|
||||
@dataclass
|
||||
class MemoryItem:
|
||||
"""记忆项（用于短期记忆）。"""
|
||||
content: str
|
||||
timestamp: datetime
|
||||
user_id: str
|
||||
importance: float = 0.5
|
||||
metadata: Dict = field(default_factory=dict)
|
||||
|
||||
def to_dict(self) -> Dict:
|
||||
"""转换为字典。"""
|
||||
return {
|
||||
'content': self.content,
|
||||
'timestamp': self.timestamp.isoformat(),
|
||||
'user_id': self.user_id,
|
||||
'importance': self.importance,
|
||||
'metadata': self.metadata
|
||||
}
|
||||
|
||||
|
||||
class ShortTermMemory:
|
||||
"""短期记忆（滑动窗口）。"""
|
||||
|
||||
def __init__(self, max_size: int = 20, max_age_minutes: int = 30):
|
||||
self.max_size = max_size
|
||||
self.max_age = timedelta(minutes=max_age_minutes)
|
||||
self.memories: Dict[str, deque] = {} # user_id -> deque of MemoryItem
|
||||
|
||||
def add(self, user_id: str, content: str, metadata: Optional[Dict] = None):
|
||||
"""添加短期记忆。"""
|
||||
if user_id not in self.memories:
|
||||
self.memories[user_id] = deque(maxlen=self.max_size)
|
||||
|
||||
memory = MemoryItem(
|
||||
content=content,
|
||||
timestamp=datetime.now(),
|
||||
user_id=user_id,
|
||||
metadata=metadata or {}
|
||||
)
|
||||
|
||||
self.memories[user_id].append(memory)
|
||||
|
||||
def get(self, user_id: str, limit: Optional[int] = None) -> List[MemoryItem]:
|
||||
"""获取短期记忆。"""
|
||||
if user_id not in self.memories:
|
||||
return []
|
||||
|
||||
# 过滤过期记忆
|
||||
now = datetime.now()
|
||||
valid_memories = [
|
||||
m for m in self.memories[user_id]
|
||||
if now - m.timestamp <= self.max_age
|
||||
]
|
||||
|
||||
if limit:
|
||||
valid_memories = valid_memories[-limit:]
|
||||
|
||||
return valid_memories
|
||||
|
||||
def clear(self, user_id: str):
|
||||
"""清除用户短期记忆。"""
|
||||
if user_id in self.memories:
|
||||
self.memories.pop(user_id, None)
|
||||
|
||||
|
||||
class MemorySystem:
|
||||
"""记忆系统：整合短期记忆与长期记忆。"""
|
||||
|
||||
def __init__(
|
||||
self,
|
||||
storage_path: Path,
|
||||
embed_func: Optional[callable] = None,
|
||||
importance_evaluator: Optional[Callable[[str, Optional[Dict]], Awaitable[float]]] = None,
|
||||
importance_threshold: float = 0.6,
|
||||
use_vector_db: bool = True
|
||||
):
|
||||
self.short_term = ShortTermMemory()
|
||||
self.embed_func = embed_func
|
||||
self.importance_evaluator = importance_evaluator
|
||||
self.importance_threshold = importance_threshold
|
||||
|
||||
# 初始化向量存储
|
||||
if use_vector_db:
|
||||
try:
|
||||
# 使用 Chroma 向量数据库
|
||||
chroma_path = storage_path.parent / "chroma_db"
|
||||
self.vector_store: VectorStore = ChromaVectorStore(chroma_path)
|
||||
logger.info("Using Chroma vector store")
|
||||
except Exception as e:
|
||||
logger.warning(f"Chroma 初始化失败，降级为 JSON 存储: {e}")
|
||||
self.vector_store = JSONVectorStore(storage_path)
|
||||
else:
|
||||
# 使用 JSON 存储（向后兼容）
|
||||
self.vector_store = JSONVectorStore(storage_path)
|
||||
logger.info("使用 JSON 存储")
|
||||
|
||||
@staticmethod
|
||||
def _normalize_embedding(values: List[float], dim: int = 1024) -> List[float]:
|
||||
if not values:
|
||||
return [0.0] * dim
|
||||
|
||||
normalized = [float(v) for v in values[:dim]]
|
||||
if len(normalized) < dim:
|
||||
normalized.extend([0.0] * (dim - len(normalized)))
|
||||
return normalized
|
||||
|
||||
@staticmethod
|
||||
def _local_embedding(text: str, dim: int = 1024) -> List[float]:
|
||||
if not text:
|
||||
return [0.0] * dim
|
||||
|
||||
vec = [0.0] * dim
|
||||
encoded = text.encode("utf-8", errors="ignore")
|
||||
if not encoded:
|
||||
return vec
|
||||
|
||||
for idx, byte in enumerate(encoded):
|
||||
bucket = idx % dim
|
||||
vec[bucket] += (byte / 255.0)
|
||||
|
||||
digest = hashlib.sha256(encoded).digest()
|
||||
for idx, byte in enumerate(digest):
|
||||
bucket = idx % dim
|
||||
vec[bucket] += ((byte / 255.0) - 0.5) * 0.1
|
||||
|
||||
return vec
|
||||
|
||||
async def _build_embedding(self, text: str) -> List[float]:
|
||||
if self.embed_func:
|
||||
try:
|
||||
embedding = await self.embed_func(text)
|
||||
if embedding:
|
||||
return [float(v) for v in list(embedding)]
|
||||
except Exception as e:
|
||||
logger.warning(f"embedding generation failed: {e}")
|
||||
|
||||
return self._local_embedding(text)
|
||||
|
||||
async def _add_vector_memory(
|
||||
self,
|
||||
memory_id: str,
|
||||
user_id: str,
|
||||
content: str,
|
||||
embedding: List[float],
|
||||
importance: float,
|
||||
metadata: Optional[Dict] = None,
|
||||
) -> bool:
|
||||
if await self.vector_store.add(
|
||||
id=memory_id,
|
||||
user_id=user_id,
|
||||
content=content,
|
||||
embedding=embedding,
|
||||
importance=importance,
|
||||
metadata=metadata,
|
||||
):
|
||||
return True
|
||||
|
||||
# Chroma collection may have a fixed historical embedding dimension.
|
||||
candidate_dims = []
|
||||
base_len = len(embedding or [])
|
||||
for dim in [base_len, 1024, 1536, 768, 384, 3072]:
|
||||
if dim and dim > 0 and dim not in candidate_dims:
|
||||
candidate_dims.append(dim)
|
||||
|
||||
for dim in candidate_dims:
|
||||
retry_embedding = self._normalize_embedding(list(embedding or []), dim=dim)
|
||||
ok = await self.vector_store.add(
|
||||
id=memory_id,
|
||||
user_id=user_id,
|
||||
content=content,
|
||||
embedding=retry_embedding,
|
||||
importance=importance,
|
||||
metadata=metadata,
|
||||
)
|
||||
if ok:
|
||||
return True
|
||||
|
||||
return False
|
||||
|
||||
@staticmethod
|
||||
def _normalize_importance(importance: float) -> float:
|
||||
try:
|
||||
value = float(importance)
|
||||
except (TypeError, ValueError):
|
||||
value = 0.5
|
||||
return max(0.0, min(1.0, value))
|
||||
|
||||
async def add_message(
|
||||
self,
|
||||
user_id: str,
|
||||
role: str,
|
||||
content: str,
|
||||
metadata: Optional[Dict] = None
|
||||
):
|
||||
"""向短期记忆添加单条消息（不做长期记忆评分）。"""
|
||||
self.short_term.add(user_id, content, metadata)
|
||||
|
||||
async def add_qa_pair(
|
||||
self,
|
||||
user_id: str,
|
||||
question: str,
|
||||
answer: str,
|
||||
metadata: Optional[Dict] = None,
|
||||
) -> Optional[VectorMemory]:
|
||||
"""
|
||||
添加最新问答对，并仅对该问答对做模型重要性评估。
|
||||
"""
|
||||
user_meta = {"role": "user"}
|
||||
assistant_meta = {"role": "assistant"}
|
||||
if isinstance(metadata, dict):
|
||||
user_meta.update(metadata)
|
||||
assistant_meta.update(metadata)
|
||||
|
||||
self.short_term.add(user_id, question, user_meta)
|
||||
self.short_term.add(user_id, answer, assistant_meta)
|
||||
|
||||
qa_content = f"用户: {question}\n助手: {answer}"
|
||||
qa_metadata = dict(metadata or {})
|
||||
qa_metadata.update({"type": "qa_pair", "question": question, "answer": answer})
|
||||
|
||||
importance = await self._evaluate_importance(qa_content, qa_metadata)
|
||||
if importance < self.importance_threshold:
|
||||
return None
|
||||
|
||||
embedding = await self._build_embedding(qa_content)
|
||||
memory_id = str(uuid.uuid4())
|
||||
ok = await self._add_vector_memory(
|
||||
memory_id=memory_id,
|
||||
user_id=user_id,
|
||||
content=qa_content,
|
||||
embedding=embedding,
|
||||
importance=importance,
|
||||
metadata=qa_metadata,
|
||||
)
|
||||
if not ok:
|
||||
return None
|
||||
|
||||
return await self.get_long_term(user_id, memory_id)
|
||||
|
||||
async def _evaluate_importance(self, content: str, metadata: Optional[Dict]) -> float:
|
||||
"""评估记忆重要性。"""
|
||||
if not content or not content.strip():
|
||||
return 0.0
|
||||
|
||||
if self.importance_evaluator:
|
||||
try:
|
||||
score = await self.importance_evaluator(content, metadata)
|
||||
return self._normalize_importance(score)
|
||||
except Exception as e:
|
||||
logger.warning(f"importance evaluation failed, fallback to neutral score: {e}")
|
||||
|
||||
# 当模型评估不可用时，使用中性分数作为兜底。
|
||||
return 0.5
|
||||
|
||||
async def get_context(
|
||||
self,
|
||||
user_id: str,
|
||||
query: Optional[str] = None,
|
||||
max_short_term: int = 10,
|
||||
max_long_term: int = 5
|
||||
) -> Tuple[List[MemoryItem], List[VectorMemory]]:
|
||||
"""获取上下文（短期 + 长期记忆）。"""
|
||||
# 获取短期记忆
|
||||
short_term_memories = self.short_term.get(user_id, limit=max_short_term)
|
||||
|
||||
# 获取相关长期记忆
|
||||
long_term_memories = []
|
||||
|
||||
if query:
|
||||
try:
|
||||
# 使用向量检索
|
||||
query_embedding = await self._build_embedding(query)
|
||||
if query_embedding:
|
||||
long_term_memories = await self.vector_store.search(
|
||||
user_id=user_id,
|
||||
query_embedding=query_embedding,
|
||||
limit=max_long_term
|
||||
)
|
||||
except Exception as e:
|
||||
logger.warning(f"向量检索失败，改用重要性检索: {e}")
|
||||
|
||||
if query and not long_term_memories:
|
||||
query_lower = query.lower()
|
||||
try:
|
||||
candidates = await self.vector_store.get_all(user_id)
|
||||
matches = [m for m in candidates if query_lower in m.content.lower()]
|
||||
matches.sort(key=lambda m: (m.importance, m.timestamp), reverse=True)
|
||||
long_term_memories = matches[:max_long_term]
|
||||
except Exception:
|
||||
pass
|
||||
|
||||
# 濡傛灉鍚戦噺妫€绱㈠け璐ユ垨娌℃湁缁撴灉锛屼娇鐢ㄩ噸瑕佹€ф绱?
|
||||
if not long_term_memories:
|
||||
long_term_memories = await self.vector_store.get_by_importance(
|
||||
user_id=user_id,
|
||||
limit=max_long_term
|
||||
)
|
||||
|
||||
# 更新长期记忆访问记录
|
||||
for memory in long_term_memories:
|
||||
await self.vector_store.update_access(memory.id)
|
||||
|
||||
return short_term_memories, long_term_memories
|
||||
|
||||
def format_context(
|
||||
self,
|
||||
short_term: List[MemoryItem],
|
||||
long_term: List[VectorMemory]
|
||||
) -> str:
|
||||
"""格式化上下文为文本。"""
|
||||
context = ""
|
||||
|
||||
if long_term:
|
||||
context += "## 相关历史记忆\n"
|
||||
for i, memory in enumerate(long_term, 1):
|
||||
context += f"{i}. {memory.content}\n"
|
||||
context += "\n"
|
||||
|
||||
if short_term:
|
||||
context += "## 最近对话\n"
|
||||
for memory in short_term:
|
||||
context += f"- {memory.content}\n"
|
||||
|
||||
return context
|
||||
|
||||
async def list_long_term(
|
||||
self, user_id: str, limit: int = 20
|
||||
) -> List[VectorMemory]:
|
||||
memories = await self.vector_store.get_all(user_id)
|
||||
memories.sort(key=lambda m: m.timestamp, reverse=True)
|
||||
if limit > 0:
|
||||
return memories[:limit]
|
||||
return memories
|
||||
|
||||
async def get_long_term(self, user_id: str, memory_id: str) -> Optional[VectorMemory]:
|
||||
memories = await self.vector_store.get_all(user_id)
|
||||
for memory in memories:
|
||||
if memory.id == memory_id:
|
||||
return memory
|
||||
return None
|
||||
|
||||
async def add_long_term(
|
||||
self,
|
||||
user_id: str,
|
||||
content: str,
|
||||
importance: float = 0.8,
|
||||
metadata: Optional[Dict] = None,
|
||||
) -> Optional[VectorMemory]:
|
||||
memory_id = str(uuid.uuid4())
|
||||
importance = self._normalize_importance(importance)
|
||||
embedding = await self._build_embedding(content)
|
||||
|
||||
ok = await self._add_vector_memory(
|
||||
memory_id=memory_id,
|
||||
user_id=user_id,
|
||||
content=content,
|
||||
embedding=embedding,
|
||||
importance=importance,
|
||||
metadata=metadata or {},
|
||||
)
|
||||
if not ok:
|
||||
return None
|
||||
|
||||
return await self.get_long_term(user_id, memory_id)
|
||||
|
||||
async def search_long_term(
|
||||
self, user_id: str, query: str, limit: int = 10
|
||||
) -> List[VectorMemory]:
|
||||
if not query:
|
||||
return []
|
||||
|
||||
query_embedding = await self._build_embedding(query)
|
||||
results = await self.vector_store.search(
|
||||
user_id=user_id,
|
||||
query_embedding=query_embedding,
|
||||
limit=limit,
|
||||
min_importance=0.0,
|
||||
)
|
||||
if results:
|
||||
return results
|
||||
|
||||
all_memories = await self.vector_store.get_all(user_id)
|
||||
query_lower = query.lower()
|
||||
matched = [m for m in all_memories if query_lower in m.content.lower()]
|
||||
matched.sort(key=lambda m: (m.importance, m.timestamp), reverse=True)
|
||||
return matched[:limit]
|
||||
|
||||
async def update_long_term(
|
||||
self,
|
||||
user_id: str,
|
||||
memory_id: str,
|
||||
content: Optional[str] = None,
|
||||
importance: Optional[float] = None,
|
||||
metadata: Optional[Dict] = None,
|
||||
) -> Optional[VectorMemory]:
|
||||
original = await self.get_long_term(user_id, memory_id)
|
||||
if not original:
|
||||
return None
|
||||
|
||||
new_content = content if content is not None else original.content
|
||||
new_importance = (
|
||||
self._normalize_importance(importance)
|
||||
if importance is not None
|
||||
else original.importance
|
||||
)
|
||||
new_metadata = dict(original.metadata or {})
|
||||
if metadata is not None:
|
||||
new_metadata = metadata
|
||||
|
||||
if content is not None:
|
||||
new_embedding = await self._build_embedding(new_content)
|
||||
else:
|
||||
new_embedding = self._normalize_embedding(list(original.embedding or []))
|
||||
|
||||
deleted = await self.vector_store.delete(memory_id)
|
||||
if not deleted:
|
||||
return None
|
||||
|
||||
added = await self._add_vector_memory(
|
||||
memory_id=memory_id,
|
||||
user_id=user_id,
|
||||
content=new_content,
|
||||
embedding=new_embedding,
|
||||
importance=new_importance,
|
||||
metadata=new_metadata,
|
||||
)
|
||||
if not added:
|
||||
return None
|
||||
|
||||
return await self.get_long_term(user_id, memory_id)
|
||||
|
||||
async def delete_long_term(self, user_id: str, memory_id: str) -> bool:
|
||||
memory = await self.get_long_term(user_id, memory_id)
|
||||
if not memory:
|
||||
return False
|
||||
return await self.vector_store.delete(memory_id)
|
||||
|
||||
def clear_short_term(self, user_id: str):
|
||||
"""清除短期记忆。"""
|
||||
self.short_term.clear(user_id)
|
||||
|
||||
async def clear_long_term(self, user_id: str):
|
||||
"""清除长期记忆。"""
|
||||
await self.vector_store.clear_user(user_id)
|
||||
|
||||
async def close(self):
|
||||
"""关闭记忆系统。"""
|
||||
await self.vector_store.close()
|
||||
|
||||
7
src/ai/models/__init__.py
Normal file
7
src/ai/models/__init__.py
Normal file
@@ -0,0 +1,7 @@
|
||||
"""
|
||||
AI模型实现
|
||||
"""
|
||||
from .openai_model import OpenAIModel
|
||||
from .anthropic_model import AnthropicModel
|
||||
|
||||
__all__ = ['OpenAIModel', 'AnthropicModel']
|
||||
120
src/ai/models/anthropic_model.py
Normal file
120
src/ai/models/anthropic_model.py
Normal file
@@ -0,0 +1,120 @@
|
||||
"""
|
||||
Anthropic Claude模型实现
|
||||
"""
|
||||
from typing import List, Optional, AsyncIterator
|
||||
from anthropic import AsyncAnthropic
|
||||
from ..base import BaseAIModel, Message, ModelConfig
|
||||
|
||||
|
||||
class AnthropicModel(BaseAIModel):
|
||||
"""Anthropic Claude模型实现"""
|
||||
|
||||
def __init__(self, config: ModelConfig):
|
||||
super().__init__(config)
|
||||
self.client = AsyncAnthropic(
|
||||
api_key=config.api_key,
|
||||
base_url=config.api_base,
|
||||
timeout=config.timeout
|
||||
)
|
||||
|
||||
async def chat(
|
||||
self,
|
||||
messages: List[Message],
|
||||
tools: Optional[List[dict]] = None,
|
||||
**kwargs
|
||||
) -> Message:
|
||||
"""同步对话"""
|
||||
# 分离system消息
|
||||
system_message = None
|
||||
formatted_messages = []
|
||||
|
||||
for msg in messages:
|
||||
if msg.role == "system":
|
||||
system_message = msg.content
|
||||
else:
|
||||
formatted_messages.append({
|
||||
"role": msg.role,
|
||||
"content": msg.content
|
||||
})
|
||||
|
||||
params = {
|
||||
"model": self.config.model_name,
|
||||
"messages": formatted_messages,
|
||||
"max_tokens": self.config.max_tokens,
|
||||
"temperature": self.config.temperature,
|
||||
}
|
||||
|
||||
if system_message:
|
||||
params["system"] = system_message
|
||||
|
||||
if tools:
|
||||
params["tools"] = tools
|
||||
|
||||
params.update(kwargs)
|
||||
|
||||
response = await self.client.messages.create(**params)
|
||||
|
||||
content = ""
|
||||
tool_calls = []
|
||||
|
||||
for block in response.content:
|
||||
if block.type == "text":
|
||||
content += block.text
|
||||
elif block.type == "tool_use":
|
||||
tool_calls.append({
|
||||
"id": block.id,
|
||||
"type": "function",
|
||||
"function": {
|
||||
"name": block.name,
|
||||
"arguments": block.input
|
||||
}
|
||||
})
|
||||
|
||||
return Message(
|
||||
role="assistant",
|
||||
content=content,
|
||||
tool_calls=tool_calls if tool_calls else None
|
||||
)
|
||||
|
||||
async def chat_stream(
|
||||
self,
|
||||
messages: List[Message],
|
||||
tools: Optional[List[dict]] = None,
|
||||
**kwargs
|
||||
) -> AsyncIterator[str]:
|
||||
"""流式对话"""
|
||||
system_message = None
|
||||
formatted_messages = []
|
||||
|
||||
for msg in messages:
|
||||
if msg.role == "system":
|
||||
system_message = msg.content
|
||||
else:
|
||||
formatted_messages.append({
|
||||
"role": msg.role,
|
||||
"content": msg.content
|
||||
})
|
||||
|
||||
params = {
|
||||
"model": self.config.model_name,
|
||||
"messages": formatted_messages,
|
||||
"max_tokens": self.config.max_tokens,
|
||||
"temperature": self.config.temperature,
|
||||
"stream": True,
|
||||
}
|
||||
|
||||
if system_message:
|
||||
params["system"] = system_message
|
||||
|
||||
if tools:
|
||||
params["tools"] = tools
|
||||
|
||||
params.update(kwargs)
|
||||
|
||||
async with self.client.messages.stream(**params) as stream:
|
||||
async for text in stream.text_stream:
|
||||
yield text
|
||||
|
||||
async def embed(self, text: str) -> List[float]:
|
||||
"""文本嵌入（Anthropic不直接提供，需要使用其他服务）"""
|
||||
raise NotImplementedError("Anthropic不提供嵌入API，请使用OpenAI或其他服务")
|
||||
235
src/ai/models/openai_model.py
Normal file
235
src/ai/models/openai_model.py
Normal file
@@ -0,0 +1,235 @@
|
||||
"""
|
||||
OpenAI模型实现（兼容OpenAI API的模型）
|
||||
"""
|
||||
import json
|
||||
import httpx
|
||||
from typing import List, Optional, AsyncIterator, Dict, Any
|
||||
from openai import AsyncOpenAI
|
||||
from ..base import BaseAIModel, Message, ModelConfig
|
||||
from src.utils.logger import setup_logger
|
||||
|
||||
logger = setup_logger('OpenAIModel')
|
||||
|
||||
|
||||
class OpenAIModel(BaseAIModel):
|
||||
"""OpenAI模型实现"""
|
||||
|
||||
def __init__(self, config: ModelConfig):
|
||||
super().__init__(config)
|
||||
self.logger = logger
|
||||
|
||||
# 创建支持UTF-8的httpx客户端
|
||||
http_client = httpx.AsyncClient(
|
||||
timeout=config.timeout,
|
||||
limits=httpx.Limits(max_keepalive_connections=5, max_connections=10)
|
||||
)
|
||||
|
||||
self.client = AsyncOpenAI(
|
||||
api_key=config.api_key,
|
||||
base_url=config.api_base,
|
||||
timeout=config.timeout,
|
||||
http_client=http_client
|
||||
)
|
||||
|
||||
async def chat(
|
||||
self,
|
||||
messages: List[Message],
|
||||
tools: Optional[List[dict]] = None,
|
||||
**kwargs
|
||||
) -> Message:
|
||||
"""同步对话"""
|
||||
formatted_messages = [self._format_message(msg) for msg in messages]
|
||||
|
||||
params = {
|
||||
"model": self.config.model_name,
|
||||
"messages": formatted_messages,
|
||||
"temperature": self.config.temperature,
|
||||
"max_tokens": self.config.max_tokens,
|
||||
"top_p": self.config.top_p,
|
||||
"frequency_penalty": self.config.frequency_penalty,
|
||||
"presence_penalty": self.config.presence_penalty,
|
||||
}
|
||||
|
||||
if tools:
|
||||
params["tools"] = tools
|
||||
|
||||
params.update(kwargs)
|
||||
|
||||
response = await self.client.chat.completions.create(**params)
|
||||
|
||||
choice = response.choices[0]
|
||||
raw_tool_calls = (
|
||||
choice.message.tool_calls
|
||||
if hasattr(choice.message, 'tool_calls') and choice.message.tool_calls
|
||||
else None
|
||||
)
|
||||
tool_calls = (
|
||||
[self._normalize_tool_call(tool_call) for tool_call in raw_tool_calls]
|
||||
if raw_tool_calls else None
|
||||
)
|
||||
return Message(
|
||||
role="assistant",
|
||||
content=choice.message.content or "",
|
||||
tool_calls=tool_calls
|
||||
)
|
||||
|
||||
async def chat_stream(
|
||||
self,
|
||||
messages: List[Message],
|
||||
tools: Optional[List[dict]] = None,
|
||||
**kwargs
|
||||
) -> AsyncIterator[str]:
|
||||
"""流式对话"""
|
||||
formatted_messages = [self._format_message(msg) for msg in messages]
|
||||
|
||||
params = {
|
||||
"model": self.config.model_name,
|
||||
"messages": formatted_messages,
|
||||
"temperature": self.config.temperature,
|
||||
"max_tokens": self.config.max_tokens,
|
||||
"stream": True,
|
||||
}
|
||||
|
||||
if tools:
|
||||
params["tools"] = tools
|
||||
|
||||
params.update(kwargs)
|
||||
|
||||
stream = await self.client.chat.completions.create(**params)
|
||||
|
||||
async for chunk in stream:
|
||||
if chunk.choices[0].delta.content:
|
||||
yield chunk.choices[0].delta.content
|
||||
|
||||
def _format_message(self, msg: Message) -> Dict[str, Any]:
|
||||
"""将内部消息结构转换为OpenAI消息格式"""
|
||||
formatted: Dict[str, Any] = {"role": msg.role}
|
||||
|
||||
if msg.role == "assistant":
|
||||
formatted["content"] = msg.content if msg.content else None
|
||||
if msg.tool_calls:
|
||||
formatted["tool_calls"] = [
|
||||
self._normalize_tool_call(tool_call)
|
||||
for tool_call in msg.tool_calls
|
||||
]
|
||||
elif msg.role == "tool":
|
||||
formatted["content"] = msg.content
|
||||
if msg.tool_call_id:
|
||||
formatted["tool_call_id"] = msg.tool_call_id
|
||||
else:
|
||||
formatted["content"] = msg.content
|
||||
if msg.name:
|
||||
formatted["name"] = msg.name
|
||||
|
||||
return formatted
|
||||
|
||||
def _normalize_tool_call(self, tool_call: Any) -> Dict[str, Any]:
|
||||
"""将工具调用对象统一转换为字典"""
|
||||
if isinstance(tool_call, dict):
|
||||
normalized = dict(tool_call)
|
||||
elif hasattr(tool_call, "model_dump"):
|
||||
normalized = tool_call.model_dump()
|
||||
else:
|
||||
function = getattr(tool_call, "function", None)
|
||||
if isinstance(function, dict):
|
||||
function_name = function.get("name")
|
||||
raw_arguments = function.get("arguments")
|
||||
else:
|
||||
function_name = getattr(function, "name", None)
|
||||
raw_arguments = getattr(function, "arguments", None)
|
||||
|
||||
normalized = {
|
||||
"id": getattr(tool_call, "id", None),
|
||||
"type": getattr(tool_call, "type", "function"),
|
||||
"function": {
|
||||
"name": function_name,
|
||||
"arguments": raw_arguments
|
||||
}
|
||||
}
|
||||
|
||||
function_data = normalized.get("function") or {}
|
||||
if not isinstance(function_data, dict):
|
||||
function_data = {
|
||||
"name": getattr(function_data, "name", None),
|
||||
"arguments": getattr(function_data, "arguments", None)
|
||||
}
|
||||
raw_arguments = function_data.get("arguments")
|
||||
|
||||
if isinstance(raw_arguments, dict):
|
||||
arguments = json.dumps(raw_arguments, ensure_ascii=False)
|
||||
elif raw_arguments is None:
|
||||
arguments = "{}"
|
||||
else:
|
||||
arguments = str(raw_arguments)
|
||||
|
||||
function_data["arguments"] = arguments
|
||||
normalized["function"] = function_data
|
||||
normalized["type"] = normalized.get("type") or "function"
|
||||
|
||||
return normalized
|
||||
|
||||
@staticmethod
|
||||
def _is_embedding_too_long_error(error: Exception) -> bool:
|
||||
status_code = getattr(error, "status_code", None)
|
||||
if status_code == 413:
|
||||
return True
|
||||
|
||||
message = str(error).lower()
|
||||
return (
|
||||
"less than 512 tokens" in message
|
||||
or "input too long" in message
|
||||
or "maximum context length" in message
|
||||
)
|
||||
|
||||
@staticmethod
|
||||
def _shrink_text_for_embedding(text: str) -> str:
|
||||
compact = " ".join((text or "").split())
|
||||
if len(compact) <= 64:
|
||||
return compact
|
||||
|
||||
target = max(64, int(len(compact) * 0.6))
|
||||
if target >= len(compact):
|
||||
target = max(64, len(compact) // 2)
|
||||
|
||||
head = target // 2
|
||||
tail = target - head
|
||||
return f"{compact[:head]} {compact[-tail:]}"
|
||||
|
||||
async def embed(self, text: str) -> List[float]:
|
||||
"""?????"""
|
||||
if isinstance(text, bytes):
|
||||
text = text.decode("utf-8", errors="ignore")
|
||||
|
||||
raw_text = str(text or "")
|
||||
candidate_text = raw_text.strip() or raw_text or " "
|
||||
retry_count = 0
|
||||
|
||||
while True:
|
||||
try:
|
||||
response = await self.client.embeddings.create(
|
||||
model=self.config.model_name,
|
||||
input=candidate_text,
|
||||
encoding_format="float"
|
||||
)
|
||||
return response.data[0].embedding
|
||||
except Exception as e:
|
||||
if self._is_embedding_too_long_error(e):
|
||||
next_text = self._shrink_text_for_embedding(candidate_text)
|
||||
if next_text and len(next_text) < len(candidate_text) and retry_count < 5:
|
||||
retry_count += 1
|
||||
self.logger.warning(
|
||||
"embedding input too long, retry with truncated text: "
|
||||
f"{len(candidate_text)} -> {len(next_text)}"
|
||||
)
|
||||
candidate_text = next_text
|
||||
continue
|
||||
|
||||
import traceback
|
||||
|
||||
self.logger.error(f"embedding generation failed: {e}")
|
||||
self.logger.error(f"model: {self.config.model_name}")
|
||||
self.logger.error(f"text length: {len(candidate_text)}")
|
||||
self.logger.error(f"text preview: {repr(candidate_text[:100])}")
|
||||
self.logger.error(f"full traceback:\n{traceback.format_exc()}")
|
||||
raise
|
||||
|
||||
232
src/ai/personality.py
Normal file
232
src/ai/personality.py
Normal file
@@ -0,0 +1,232 @@
|
||||
"""Personality system for role-play profiles."""
|
||||
|
||||
from dataclasses import dataclass, field
|
||||
from enum import Enum
|
||||
import json
|
||||
from pathlib import Path
|
||||
from typing import Dict, List, Optional
|
||||
|
||||
|
||||
class PersonalityTrait(Enum):
|
||||
"""Personality traits."""
|
||||
|
||||
FRIENDLY = "Friendly"
|
||||
PROFESSIONAL = "Professional"
|
||||
HUMOROUS = "Humorous"
|
||||
SERIOUS = "Serious"
|
||||
CREATIVE = "Creative"
|
||||
ANALYTICAL = "Analytical"
|
||||
EMPATHETIC = "Empathetic"
|
||||
DIRECT = "Direct"
|
||||
|
||||
|
||||
@dataclass
|
||||
class PersonalityProfile:
|
||||
"""Single personality profile."""
|
||||
|
||||
name: str
|
||||
description: str
|
||||
traits: List[PersonalityTrait]
|
||||
speaking_style: str
|
||||
example_responses: List[str] = field(default_factory=list)
|
||||
custom_instructions: str = ""
|
||||
|
||||
def to_system_prompt(self) -> str:
|
||||
"""Build plain-text system prompt."""
|
||||
|
||||
traits_text = ", ".join([t.value for t in self.traits]) if self.traits else "Friendly"
|
||||
lines = [
|
||||
"Role Setting",
|
||||
f"You are {self.name}. {self.description}",
|
||||
f"Traits: {traits_text}",
|
||||
f"Speaking style: {self.speaking_style}",
|
||||
"Output rule: plain text only. Do not use Markdown syntax.",
|
||||
]
|
||||
|
||||
if self.example_responses:
|
||||
lines.append("Reference responses:")
|
||||
for idx, example in enumerate(self.example_responses, 1):
|
||||
lines.append(f"{idx}. {example}")
|
||||
|
||||
if self.custom_instructions:
|
||||
lines.append(f"Additional instructions: {self.custom_instructions}")
|
||||
|
||||
return "\n".join(lines)
|
||||
|
||||
|
||||
class PersonalitySystem:
|
||||
"""Personality management and persistence."""
|
||||
|
||||
def __init__(self, config_path: Optional[Path] = None):
|
||||
self.config_path = config_path or Path("config/personalities.json")
|
||||
self.personalities: Dict[str, PersonalityProfile] = {}
|
||||
self.current_personality: Optional[PersonalityProfile] = None
|
||||
self._load_personalities()
|
||||
|
||||
def _dict_to_profile(self, config: Dict) -> PersonalityProfile:
|
||||
trait_names = config.get("traits", [])
|
||||
traits: List[PersonalityTrait] = []
|
||||
|
||||
for trait_name in trait_names:
|
||||
if trait_name in PersonalityTrait.__members__:
|
||||
traits.append(PersonalityTrait[trait_name])
|
||||
|
||||
if not traits:
|
||||
traits = [PersonalityTrait.FRIENDLY]
|
||||
|
||||
return PersonalityProfile(
|
||||
name=str(config.get("name", "Assistant")),
|
||||
description=str(config.get("description", "")),
|
||||
traits=traits,
|
||||
speaking_style=str(config.get("speaking_style", "Natural and concise")),
|
||||
example_responses=list(config.get("example_responses", [])),
|
||||
custom_instructions=str(config.get("custom_instructions", "")),
|
||||
)
|
||||
|
||||
def _load_personalities(self):
|
||||
"""Load personality config from disk or create defaults."""
|
||||
|
||||
if self.config_path.exists():
|
||||
with open(self.config_path, "r", encoding="utf-8") as f:
|
||||
data = json.load(f)
|
||||
for key, config in data.items():
|
||||
self.personalities[key] = self._dict_to_profile(config)
|
||||
|
||||
if "default" in self.personalities:
|
||||
self.current_personality = self.personalities["default"]
|
||||
elif self.personalities:
|
||||
first_key = next(iter(self.personalities.keys()))
|
||||
self.current_personality = self.personalities[first_key]
|
||||
return
|
||||
|
||||
self._create_default_personalities()
|
||||
|
||||
def _create_default_personalities(self):
|
||||
"""Create and persist built-in default profiles."""
|
||||
|
||||
default = PersonalityProfile(
|
||||
name="Assistant",
|
||||
description="A friendly and practical AI assistant.",
|
||||
traits=[
|
||||
PersonalityTrait.FRIENDLY,
|
||||
PersonalityTrait.PROFESSIONAL,
|
||||
PersonalityTrait.EMPATHETIC,
|
||||
],
|
||||
speaking_style="Warm, clear, and actionable.",
|
||||
example_responses=[
|
||||
"I understand. Let's solve this step by step.",
|
||||
"Here is the result first, then the key details.",
|
||||
],
|
||||
)
|
||||
|
||||
tech_expert = PersonalityProfile(
|
||||
name="Tech Expert",
|
||||
description="A senior engineer focused on correctness and maintainability.",
|
||||
traits=[
|
||||
PersonalityTrait.PROFESSIONAL,
|
||||
PersonalityTrait.ANALYTICAL,
|
||||
PersonalityTrait.DIRECT,
|
||||
],
|
||||
speaking_style="Direct, structured, and implementation-oriented.",
|
||||
)
|
||||
|
||||
creative = PersonalityProfile(
|
||||
name="Creative Partner",
|
||||
description="A collaborative creative role-play partner.",
|
||||
traits=[
|
||||
PersonalityTrait.CREATIVE,
|
||||
PersonalityTrait.HUMOROUS,
|
||||
PersonalityTrait.FRIENDLY,
|
||||
],
|
||||
speaking_style="Lively, imaginative, and expressive.",
|
||||
)
|
||||
|
||||
self.personalities = {
|
||||
"default": default,
|
||||
"tech_expert": tech_expert,
|
||||
"creative": creative,
|
||||
}
|
||||
self.current_personality = default
|
||||
self._save_personalities()
|
||||
|
||||
def _save_personalities(self):
|
||||
"""Persist personalities to disk."""
|
||||
|
||||
self.config_path.parent.mkdir(parents=True, exist_ok=True)
|
||||
data = {}
|
||||
|
||||
for key, profile in self.personalities.items():
|
||||
data[key] = {
|
||||
"name": profile.name,
|
||||
"description": profile.description,
|
||||
"traits": [trait.name for trait in profile.traits],
|
||||
"speaking_style": profile.speaking_style,
|
||||
"example_responses": profile.example_responses,
|
||||
"custom_instructions": profile.custom_instructions,
|
||||
}
|
||||
|
||||
with open(self.config_path, "w", encoding="utf-8") as f:
|
||||
json.dump(data, f, ensure_ascii=False, indent=2)
|
||||
|
||||
def set_personality(self, key: str) -> bool:
|
||||
"""Switch active personality by key."""
|
||||
|
||||
if key not in self.personalities:
|
||||
return False
|
||||
|
||||
self.current_personality = self.personalities[key]
|
||||
return True
|
||||
|
||||
def get_system_prompt(self) -> str:
|
||||
"""Get current personality prompt."""
|
||||
|
||||
if self.current_personality:
|
||||
return self.current_personality.to_system_prompt()
|
||||
return ""
|
||||
|
||||
def add_personality(self, key: str, profile: PersonalityProfile) -> bool:
|
||||
"""Add a new personality profile."""
|
||||
|
||||
key = key.strip()
|
||||
if not key:
|
||||
return False
|
||||
|
||||
self.personalities[key] = profile
|
||||
if not self.current_personality:
|
||||
self.current_personality = profile
|
||||
self._save_personalities()
|
||||
return True
|
||||
|
||||
def remove_personality(self, key: str) -> bool:
|
||||
"""Remove a personality profile."""
|
||||
|
||||
if key == "default":
|
||||
return False
|
||||
|
||||
if key not in self.personalities:
|
||||
return False
|
||||
|
||||
removed_profile = self.personalities[key]
|
||||
del self.personalities[key]
|
||||
|
||||
if self.current_personality == removed_profile:
|
||||
if "default" in self.personalities:
|
||||
self.current_personality = self.personalities["default"]
|
||||
elif self.personalities:
|
||||
first_key = next(iter(self.personalities.keys()))
|
||||
self.current_personality = self.personalities[first_key]
|
||||
else:
|
||||
self.current_personality = None
|
||||
|
||||
self._save_personalities()
|
||||
return True
|
||||
|
||||
def list_personalities(self) -> List[str]:
|
||||
"""List all personality keys."""
|
||||
|
||||
return sorted(self.personalities.keys())
|
||||
|
||||
def get_personality(self, key: str) -> Optional[PersonalityProfile]:
|
||||
"""Get personality by key."""
|
||||
|
||||
return self.personalities.get(key)
|
||||
6
src/ai/skills/__init__.py
Normal file
6
src/ai/skills/__init__.py
Normal file
@@ -0,0 +1,6 @@
|
||||
"""
|
||||
Skills系统初始化
|
||||
"""
|
||||
from .base import Skill, SkillsManager, SkillMetadata, create_skill_template
|
||||
|
||||
__all__ = ['Skill', 'SkillsManager', 'SkillMetadata', 'create_skill_template']
|
||||
552
src/ai/skills/base.py
Normal file
552
src/ai/skills/base.py
Normal file
@@ -0,0 +1,552 @@
|
||||
"""
|
||||
Skills 系统 - 可扩展技能插件框架。
|
||||
"""
|
||||
|
||||
from dataclasses import dataclass
|
||||
import importlib
|
||||
import inspect
|
||||
import json
|
||||
from pathlib import Path
|
||||
import re
|
||||
import shutil
|
||||
import sys
|
||||
import tempfile
|
||||
import time
|
||||
from typing import Any, Callable, Dict, List, Optional, Tuple
|
||||
import urllib.request
|
||||
import zipfile
|
||||
import os
|
||||
import stat
|
||||
|
||||
from src.utils.logger import setup_logger
|
||||
|
||||
logger = setup_logger("SkillsSystem")
|
||||
|
||||
|
||||
@dataclass
|
||||
class SkillMetadata:
|
||||
"""技能元数据。"""
|
||||
|
||||
name: str
|
||||
version: str
|
||||
description: str
|
||||
author: str
|
||||
dependencies: List[str]
|
||||
enabled: bool = True
|
||||
|
||||
|
||||
class Skill:
|
||||
"""技能基类。"""
|
||||
|
||||
def __init__(self):
|
||||
self.metadata: Optional[SkillMetadata] = None
|
||||
self.tools: Dict[str, Callable] = {}
|
||||
self.manager = None
|
||||
|
||||
async def initialize(self):
|
||||
"""初始化技能。"""
|
||||
|
||||
async def cleanup(self):
|
||||
"""清理技能。"""
|
||||
|
||||
def get_tools(self) -> Dict[str, Callable]:
|
||||
"""获取技能提供的工具。"""
|
||||
|
||||
return self.tools
|
||||
|
||||
def register_tool(self, name: str, func: Callable):
|
||||
"""注册工具。"""
|
||||
|
||||
self.tools[name] = func
|
||||
|
||||
|
||||
class SkillsManager:
|
||||
"""技能管理器。"""
|
||||
|
||||
_SKILL_KEY_PATTERN = re.compile(r"[^a-zA-Z0-9_]")
|
||||
_GITHUB_SHORTCUT_PATTERN = re.compile(
|
||||
r"^[A-Za-z0-9_.-]+/[A-Za-z0-9_.-]+(?:#[A-Za-z0-9_.-]+)?$"
|
||||
)
|
||||
|
||||
def __init__(self, skills_dir: Path):
|
||||
self.skills_dir = skills_dir
|
||||
self.skills: Dict[str, Skill] = {}
|
||||
self.skills_dir.mkdir(parents=True, exist_ok=True)
|
||||
logger.info(f"✅ Skills 目录: {skills_dir}")
|
||||
|
||||
@classmethod
|
||||
def normalize_skill_key(cls, raw_name: str) -> str:
|
||||
"""将任意输入规范化为可导入的 Python 包名。"""
|
||||
|
||||
key = raw_name.strip().lower().replace("-", "_").replace(" ", "_")
|
||||
key = cls._SKILL_KEY_PATTERN.sub("_", key)
|
||||
key = re.sub(r"_+", "_", key).strip("_")
|
||||
|
||||
if not key:
|
||||
raise ValueError("技能名不能为空")
|
||||
|
||||
if key[0].isdigit():
|
||||
key = f"skill_{key}"
|
||||
|
||||
return key
|
||||
|
||||
def _get_skill_path(self, skill_name: str) -> Path:
|
||||
return self.skills_dir / self.normalize_skill_key(skill_name)
|
||||
|
||||
@staticmethod
|
||||
def _on_rmtree_error(func, path, exc_info):
|
||||
"""Handle Windows readonly/locked file deletion errors."""
|
||||
try:
|
||||
os.chmod(path, stat.S_IWRITE)
|
||||
func(path)
|
||||
except Exception:
|
||||
# Keep original failure path for upper retry logic.
|
||||
pass
|
||||
|
||||
def _read_metadata(self, skill_path: Path, fallback_name: str) -> Dict[str, Any]:
|
||||
metadata_file = skill_path / "skill.json"
|
||||
if metadata_file.exists():
|
||||
with open(metadata_file, "r", encoding="utf-8") as f:
|
||||
metadata = json.load(f)
|
||||
else:
|
||||
metadata = {}
|
||||
|
||||
metadata.setdefault("name", fallback_name)
|
||||
metadata.setdefault("version", "1.0.0")
|
||||
metadata.setdefault("description", f"{fallback_name} skill")
|
||||
metadata.setdefault("author", "unknown")
|
||||
metadata.setdefault("dependencies", [])
|
||||
metadata.setdefault("enabled", True)
|
||||
|
||||
with open(metadata_file, "w", encoding="utf-8") as f:
|
||||
json.dump(metadata, f, ensure_ascii=False, indent=2)
|
||||
|
||||
return metadata
|
||||
|
||||
def _ensure_skill_package_layout(self, skill_path: Path, skill_key: str):
|
||||
"""确保技能目录满足运行最小结构。"""
|
||||
|
||||
skill_path.mkdir(parents=True, exist_ok=True)
|
||||
|
||||
init_file = skill_path / "__init__.py"
|
||||
if not init_file.exists():
|
||||
init_file.write_text("", encoding="utf-8")
|
||||
|
||||
main_file = skill_path / "main.py"
|
||||
if not main_file.exists():
|
||||
template = f'''"""{skill_key} skill"""
|
||||
from src.ai.skills.base import Skill
|
||||
|
||||
|
||||
class {"".join(p.capitalize() for p in skill_key.split("_"))}Skill(Skill):
|
||||
async def initialize(self):
|
||||
self.register_tool("ping", self.ping)
|
||||
|
||||
async def ping(self, text: str = "ok") -> str:
|
||||
return text
|
||||
|
||||
async def cleanup(self):
|
||||
pass
|
||||
'''
|
||||
main_file.write_text(template, encoding="utf-8")
|
||||
|
||||
self._read_metadata(skill_path, skill_key)
|
||||
|
||||
async def load_skill(self, skill_name: str) -> bool:
|
||||
"""加载技能。"""
|
||||
|
||||
try:
|
||||
skill_name = self.normalize_skill_key(skill_name)
|
||||
|
||||
if skill_name in self.skills:
|
||||
logger.info(f"✅ 技能已加载: {skill_name}")
|
||||
return True
|
||||
|
||||
skill_path = self._get_skill_path(skill_name)
|
||||
if not skill_path.exists():
|
||||
logger.error(f"❌ 技能不存在: {skill_name}")
|
||||
return False
|
||||
|
||||
metadata_file = skill_path / "skill.json"
|
||||
if not metadata_file.exists():
|
||||
logger.error(f"❌ 技能元数据不存在: {skill_name}")
|
||||
return False
|
||||
|
||||
with open(metadata_file, "r", encoding="utf-8") as f:
|
||||
metadata_dict = json.load(f)
|
||||
|
||||
metadata = SkillMetadata(**metadata_dict)
|
||||
|
||||
if not metadata.enabled:
|
||||
logger.info(f"⏸️ 技能已禁用: {skill_name}")
|
||||
return False
|
||||
|
||||
module_path = f"skills.{skill_name}.main"
|
||||
importlib.invalidate_caches()
|
||||
|
||||
try:
|
||||
old_dont_write = sys.dont_write_bytecode
|
||||
sys.dont_write_bytecode = True
|
||||
try:
|
||||
if module_path in sys.modules:
|
||||
module = importlib.reload(sys.modules[module_path])
|
||||
else:
|
||||
module = importlib.import_module(module_path)
|
||||
finally:
|
||||
sys.dont_write_bytecode = old_dont_write
|
||||
except Exception as exc:
|
||||
logger.error(f"❌ 无法导入技能模块 {module_path}: {exc}")
|
||||
return False
|
||||
|
||||
skill_class = None
|
||||
for _, obj in inspect.getmembers(module):
|
||||
if inspect.isclass(obj) and issubclass(obj, Skill) and obj != Skill:
|
||||
skill_class = obj
|
||||
break
|
||||
|
||||
if not skill_class:
|
||||
logger.error(f"❌ 技能中未找到 Skill 子类: {skill_name}")
|
||||
return False
|
||||
|
||||
skill = skill_class()
|
||||
skill.metadata = metadata
|
||||
skill.manager = self
|
||||
|
||||
await skill.initialize()
|
||||
self.skills[skill_name] = skill
|
||||
|
||||
logger.info(f"✅ 加载技能: {skill_name} v{metadata.version}")
|
||||
return True
|
||||
|
||||
except Exception as exc:
|
||||
logger.error(f"❌ 加载技能失败 {skill_name}: {exc}")
|
||||
return False
|
||||
|
||||
async def load_all_skills(self):
|
||||
"""加载所有可用技能。"""
|
||||
|
||||
for skill_name in self.list_available_skills():
|
||||
await self.load_skill(skill_name)
|
||||
|
||||
async def unload_skill(self, skill_name: str) -> bool:
|
||||
"""仅卸载内存中的技能。"""
|
||||
|
||||
skill_name = self.normalize_skill_key(skill_name)
|
||||
if skill_name not in self.skills:
|
||||
return False
|
||||
|
||||
skill = self.skills[skill_name]
|
||||
await skill.cleanup()
|
||||
del self.skills[skill_name]
|
||||
|
||||
sys.modules.pop(f"skills.{skill_name}.main", None)
|
||||
sys.modules.pop(f"skills.{skill_name}", None)
|
||||
importlib.invalidate_caches()
|
||||
|
||||
logger.info(f"✅ 卸载技能: {skill_name}")
|
||||
return True
|
||||
|
||||
async def uninstall_skill(self, skill_name: str, delete_files: bool = True) -> bool:
|
||||
"""卸载技能并可选删除文件。"""
|
||||
|
||||
skill_name = self.normalize_skill_key(skill_name)
|
||||
|
||||
if skill_name in self.skills:
|
||||
await self.unload_skill(skill_name)
|
||||
|
||||
if not delete_files:
|
||||
return True
|
||||
|
||||
skill_path = self._get_skill_path(skill_name)
|
||||
if not skill_path.exists():
|
||||
return False
|
||||
|
||||
removed = False
|
||||
for _ in range(3):
|
||||
try:
|
||||
shutil.rmtree(skill_path, ignore_errors=False, onerror=self._on_rmtree_error)
|
||||
except PermissionError:
|
||||
pass
|
||||
|
||||
if not skill_path.exists():
|
||||
removed = True
|
||||
break
|
||||
|
||||
time.sleep(0.2)
|
||||
|
||||
if not removed:
|
||||
try:
|
||||
metadata_file = skill_path / "skill.json"
|
||||
metadata = {}
|
||||
if metadata_file.exists():
|
||||
with open(metadata_file, "r", encoding="utf-8") as f:
|
||||
metadata = json.load(f)
|
||||
metadata["enabled"] = False
|
||||
with open(metadata_file, "w", encoding="utf-8") as f:
|
||||
json.dump(metadata, f, ensure_ascii=False, indent=2)
|
||||
logger.warning(f"⚠️ 删除目录失败，已软卸载技能: {skill_name}")
|
||||
return True
|
||||
except Exception:
|
||||
return False
|
||||
|
||||
importlib.invalidate_caches()
|
||||
logger.info(f"✅ 删除技能目录: {skill_name}")
|
||||
return True
|
||||
|
||||
def get_skill(self, skill_name: str) -> Optional[Skill]:
|
||||
"""获取已加载技能实例。"""
|
||||
|
||||
skill_name = self.normalize_skill_key(skill_name)
|
||||
return self.skills.get(skill_name)
|
||||
|
||||
def list_skills(self) -> List[str]:
|
||||
"""列出已加载技能。"""
|
||||
|
||||
return sorted(self.skills.keys())
|
||||
|
||||
def list_available_skills(self) -> List[str]:
|
||||
"""列出可加载技能目录。"""
|
||||
|
||||
if not self.skills_dir.exists():
|
||||
return []
|
||||
|
||||
available: List[str] = []
|
||||
for skill_dir in self.skills_dir.iterdir():
|
||||
if not skill_dir.is_dir() or skill_dir.name.startswith("_"):
|
||||
continue
|
||||
|
||||
if (skill_dir / "skill.json").exists() and (skill_dir / "main.py").exists():
|
||||
try:
|
||||
with open(skill_dir / "skill.json", "r", encoding="utf-8") as f:
|
||||
metadata = json.load(f)
|
||||
if not metadata.get("enabled", True):
|
||||
continue
|
||||
available.append(self.normalize_skill_key(skill_dir.name))
|
||||
except ValueError:
|
||||
continue
|
||||
except Exception:
|
||||
continue
|
||||
|
||||
return sorted(set(available))
|
||||
|
||||
def get_all_tools(self) -> Dict[str, Callable]:
|
||||
"""获取全部技能工具。"""
|
||||
|
||||
all_tools: Dict[str, Callable] = {}
|
||||
for skill_name, skill in self.skills.items():
|
||||
for tool_name, tool_func in skill.get_tools().items():
|
||||
all_tools[f"{skill_name}.{tool_name}"] = tool_func
|
||||
return all_tools
|
||||
|
||||
async def reload_skill(self, skill_name: str) -> bool:
|
||||
"""重载技能。"""
|
||||
|
||||
skill_name = self.normalize_skill_key(skill_name)
|
||||
if skill_name in self.skills:
|
||||
await self.unload_skill(skill_name)
|
||||
return await self.load_skill(skill_name)
|
||||
|
||||
def _resolve_network_url(self, source: str) -> str:
|
||||
"""支持 URL 与 GitHub 简写。"""
|
||||
|
||||
source = source.strip()
|
||||
if source.startswith(("http://", "https://")):
|
||||
return source
|
||||
|
||||
if self._GITHUB_SHORTCUT_PATTERN.match(source):
|
||||
repo, _, branch = source.partition("#")
|
||||
branch = branch or "main"
|
||||
return f"https://codeload.github.com/{repo}/zip/refs/heads/{branch}"
|
||||
|
||||
raise ValueError("source 必须是 URL 或 owner/repo[#branch]")
|
||||
|
||||
def _download_zip(self, url: str, output_zip: Path):
|
||||
"""下载 zip 包到本地。"""
|
||||
|
||||
req = urllib.request.Request(url, headers={"User-Agent": "QQBot-Skills/1.0"})
|
||||
with urllib.request.urlopen(req, timeout=30) as resp:
|
||||
data = resp.read()
|
||||
output_zip.write_bytes(data)
|
||||
|
||||
def _find_skill_candidates(self, root_dir: Path) -> List[Tuple[str, Path]]:
|
||||
"""在目录中扫描技能候选项。"""
|
||||
|
||||
candidates: List[Tuple[str, Path]] = []
|
||||
for metadata_file in root_dir.rglob("skill.json"):
|
||||
candidate_dir = metadata_file.parent
|
||||
if not (candidate_dir / "main.py").exists():
|
||||
continue
|
||||
|
||||
try:
|
||||
with open(metadata_file, "r", encoding="utf-8") as f:
|
||||
metadata = json.load(f)
|
||||
raw_name = str(metadata.get("name") or candidate_dir.name)
|
||||
except Exception:
|
||||
raw_name = candidate_dir.name
|
||||
|
||||
try:
|
||||
skill_key = self.normalize_skill_key(raw_name)
|
||||
except ValueError:
|
||||
continue
|
||||
|
||||
candidates.append((skill_key, candidate_dir))
|
||||
|
||||
uniq: Dict[str, Path] = {}
|
||||
for key, path in candidates:
|
||||
uniq[key] = path
|
||||
|
||||
return sorted(uniq.items(), key=lambda x: x[0])
|
||||
|
||||
def install_skill_from_source(
|
||||
self,
|
||||
source: str,
|
||||
skill_name: Optional[str] = None,
|
||||
overwrite: bool = False,
|
||||
) -> Tuple[bool, str]:
|
||||
"""从网络或本地源安装技能目录（仅落盘，不自动加载）。"""
|
||||
|
||||
desired_key = self.normalize_skill_key(skill_name) if skill_name else None
|
||||
|
||||
with tempfile.TemporaryDirectory(prefix="qqbot_skill_") as tmp:
|
||||
tmp_dir = Path(tmp)
|
||||
extract_root: Optional[Path] = None
|
||||
|
||||
source_path = Path(source)
|
||||
if source_path.exists():
|
||||
if source_path.is_dir():
|
||||
extract_root = source_path
|
||||
elif source_path.is_file() and source_path.suffix.lower() == ".zip":
|
||||
with zipfile.ZipFile(source_path, "r") as zf:
|
||||
zf.extractall(tmp_dir / "extract")
|
||||
extract_root = tmp_dir / "extract"
|
||||
else:
|
||||
return False, "本地 source 仅支持目录或 zip 文件"
|
||||
else:
|
||||
try:
|
||||
url = self._resolve_network_url(source)
|
||||
except ValueError as exc:
|
||||
return False, str(exc)
|
||||
|
||||
download_zip = tmp_dir / "download.zip"
|
||||
try:
|
||||
self._download_zip(url, download_zip)
|
||||
except Exception as exc:
|
||||
# GitHub 简写默认 main 失败时尝试 master
|
||||
if "codeload.github.com" in url and url.endswith("/main"):
|
||||
fallback = url[:-4] + "master"
|
||||
try:
|
||||
self._download_zip(fallback, download_zip)
|
||||
except Exception:
|
||||
return False, f"下载技能失败: {exc}"
|
||||
else:
|
||||
return False, f"下载技能失败: {exc}"
|
||||
|
||||
try:
|
||||
with zipfile.ZipFile(download_zip, "r") as zf:
|
||||
zf.extractall(tmp_dir / "extract")
|
||||
except Exception as exc:
|
||||
return False, f"解压技能失败: {exc}"
|
||||
|
||||
extract_root = tmp_dir / "extract"
|
||||
|
||||
candidates = self._find_skill_candidates(extract_root)
|
||||
if not candidates:
|
||||
return False, "未找到可安装技能（需包含 skill.json 与 main.py）"
|
||||
|
||||
selected_key: Optional[str] = None
|
||||
selected_path: Optional[Path] = None
|
||||
|
||||
if desired_key:
|
||||
for key, path in candidates:
|
||||
if key == desired_key:
|
||||
selected_key, selected_path = key, path
|
||||
break
|
||||
|
||||
if not selected_path:
|
||||
names = ", ".join([k for k, _ in candidates])
|
||||
return False, f"源中未找到技能 {desired_key}，可选: {names}"
|
||||
else:
|
||||
if len(candidates) > 1:
|
||||
names = ", ".join([k for k, _ in candidates])
|
||||
return False, f"检测到多个技能，请指定 skill_name。可选: {names}"
|
||||
selected_key, selected_path = candidates[0]
|
||||
|
||||
assert selected_key is not None and selected_path is not None
|
||||
target_path = self._get_skill_path(selected_key)
|
||||
|
||||
if target_path.exists():
|
||||
if not overwrite:
|
||||
return False, f"技能已存在: {selected_key}"
|
||||
shutil.rmtree(target_path)
|
||||
|
||||
shutil.copytree(
|
||||
selected_path,
|
||||
target_path,
|
||||
ignore=shutil.ignore_patterns("__pycache__", "*.pyc", ".git", ".github"),
|
||||
)
|
||||
self._ensure_skill_package_layout(target_path, selected_key)
|
||||
importlib.invalidate_caches()
|
||||
|
||||
logger.info(f"✅ 安装技能成功: {selected_key} <- {source}")
|
||||
return True, selected_key
|
||||
|
||||
|
||||
def create_skill_template(
|
||||
skill_name: str,
|
||||
output_dir: Path,
|
||||
description: str = "技能描述",
|
||||
author: str = "QQBot",
|
||||
):
|
||||
"""创建技能模板。"""
|
||||
|
||||
skill_key = SkillsManager.normalize_skill_key(skill_name)
|
||||
skill_dir = output_dir / skill_key
|
||||
skill_dir.mkdir(parents=True, exist_ok=True)
|
||||
|
||||
metadata = {
|
||||
"name": skill_key,
|
||||
"version": "1.0.0",
|
||||
"description": description,
|
||||
"author": author,
|
||||
"dependencies": [],
|
||||
"enabled": True,
|
||||
}
|
||||
|
||||
with open(skill_dir / "skill.json", "w", encoding="utf-8") as f:
|
||||
json.dump(metadata, f, ensure_ascii=False, indent=2)
|
||||
|
||||
class_name = "".join(word.capitalize() for word in skill_key.split("_")) + "Skill"
|
||||
main_code = f'''"""{skill_key} skill"""
|
||||
from src.ai.skills.base import Skill
|
||||
|
||||
|
||||
class {class_name}(Skill):
|
||||
async def initialize(self):
|
||||
self.register_tool("example_tool", self.example_tool)
|
||||
|
||||
async def example_tool(self, text: str) -> str:
|
||||
return f"{skill_key} 收到: {{text}}"
|
||||
|
||||
async def cleanup(self):
|
||||
pass
|
||||
'''
|
||||
|
||||
with open(skill_dir / "main.py", "w", encoding="utf-8") as f:
|
||||
f.write(main_code)
|
||||
|
||||
with open(skill_dir / "__init__.py", "w", encoding="utf-8") as f:
|
||||
f.write("")
|
||||
|
||||
readme = f"""# {skill_key}
|
||||
|
||||
## 描述
|
||||
{description}
|
||||
|
||||
## 工具
|
||||
- example_tool(text)
|
||||
"""
|
||||
|
||||
with open(skill_dir / "README.md", "w", encoding="utf-8") as f:
|
||||
f.write(readme)
|
||||
|
||||
logger.info(f"✅ 创建技能模板: {skill_dir}")
|
||||
327
src/ai/task_manager.py
Normal file
327
src/ai/task_manager.py
Normal file
@@ -0,0 +1,327 @@
|
||||
"""
|
||||
长任务管理器 - 处理需要多步骤的复杂任务
|
||||
"""
|
||||
import asyncio
|
||||
from typing import List, Dict, Optional, Callable, Any
|
||||
from dataclasses import dataclass, field
|
||||
from datetime import datetime
|
||||
from enum import Enum
|
||||
import json
|
||||
from pathlib import Path
|
||||
import uuid
|
||||
|
||||
|
||||
class TaskStatus(Enum):
|
||||
"""任务状态"""
|
||||
PENDING = "pending"
|
||||
RUNNING = "running"
|
||||
PAUSED = "paused"
|
||||
COMPLETED = "completed"
|
||||
FAILED = "failed"
|
||||
CANCELLED = "cancelled"
|
||||
|
||||
|
||||
@dataclass
|
||||
class TaskStep:
|
||||
"""任务步骤"""
|
||||
step_id: str
|
||||
description: str
|
||||
action: str
|
||||
params: Dict[str, Any]
|
||||
status: TaskStatus = TaskStatus.PENDING
|
||||
result: Optional[Any] = None
|
||||
error: Optional[str] = None
|
||||
started_at: Optional[datetime] = None
|
||||
completed_at: Optional[datetime] = None
|
||||
|
||||
def to_dict(self) -> Dict:
|
||||
return {
|
||||
'step_id': self.step_id,
|
||||
'description': self.description,
|
||||
'action': self.action,
|
||||
'params': self.params,
|
||||
'status': self.status.value,
|
||||
'result': self.result,
|
||||
'error': self.error,
|
||||
'started_at': self.started_at.isoformat() if self.started_at else None,
|
||||
'completed_at': self.completed_at.isoformat() if self.completed_at else None
|
||||
}
|
||||
|
||||
|
||||
@dataclass
|
||||
class LongTask:
|
||||
"""长任务"""
|
||||
task_id: str
|
||||
user_id: str
|
||||
title: str
|
||||
description: str
|
||||
steps: List[TaskStep] = field(default_factory=list)
|
||||
status: TaskStatus = TaskStatus.PENDING
|
||||
created_at: datetime = field(default_factory=datetime.now)
|
||||
started_at: Optional[datetime] = None
|
||||
completed_at: Optional[datetime] = None
|
||||
progress: float = 0.0
|
||||
metadata: Dict = field(default_factory=dict)
|
||||
|
||||
def to_dict(self) -> Dict:
|
||||
return {
|
||||
'task_id': self.task_id,
|
||||
'user_id': self.user_id,
|
||||
'title': self.title,
|
||||
'description': self.description,
|
||||
'steps': [step.to_dict() for step in self.steps],
|
||||
'status': self.status.value,
|
||||
'created_at': self.created_at.isoformat(),
|
||||
'started_at': self.started_at.isoformat() if self.started_at else None,
|
||||
'completed_at': self.completed_at.isoformat() if self.completed_at else None,
|
||||
'progress': self.progress,
|
||||
'metadata': self.metadata
|
||||
}
|
||||
|
||||
@classmethod
|
||||
def from_dict(cls, data: Dict) -> 'LongTask':
|
||||
steps = [
|
||||
TaskStep(
|
||||
step_id=s['step_id'],
|
||||
description=s['description'],
|
||||
action=s['action'],
|
||||
params=s['params'],
|
||||
status=TaskStatus(s['status']),
|
||||
result=s.get('result'),
|
||||
error=s.get('error'),
|
||||
started_at=datetime.fromisoformat(s['started_at']) if s.get('started_at') else None,
|
||||
completed_at=datetime.fromisoformat(s['completed_at']) if s.get('completed_at') else None
|
||||
)
|
||||
for s in data['steps']
|
||||
]
|
||||
|
||||
return cls(
|
||||
task_id=data['task_id'],
|
||||
user_id=data['user_id'],
|
||||
title=data['title'],
|
||||
description=data['description'],
|
||||
steps=steps,
|
||||
status=TaskStatus(data['status']),
|
||||
created_at=datetime.fromisoformat(data['created_at']),
|
||||
started_at=datetime.fromisoformat(data['started_at']) if data.get('started_at') else None,
|
||||
completed_at=datetime.fromisoformat(data['completed_at']) if data.get('completed_at') else None,
|
||||
progress=data.get('progress', 0.0),
|
||||
metadata=data.get('metadata', {})
|
||||
)
|
||||
|
||||
|
||||
class LongTaskManager:
|
||||
"""长任务管理器"""
|
||||
|
||||
def __init__(self, storage_path: Path):
|
||||
self.storage_path = storage_path
|
||||
self.tasks: Dict[str, LongTask] = {}
|
||||
self.action_handlers: Dict[str, Callable] = {}
|
||||
self.running_tasks: Dict[str, asyncio.Task] = {}
|
||||
self._load()
|
||||
|
||||
def _load(self):
|
||||
"""加载任务"""
|
||||
if self.storage_path.exists():
|
||||
with open(self.storage_path, 'r', encoding='utf-8') as f:
|
||||
data = json.load(f)
|
||||
for task_data in data:
|
||||
task = LongTask.from_dict(task_data)
|
||||
self.tasks[task.task_id] = task
|
||||
|
||||
def _save(self):
|
||||
"""保存任务"""
|
||||
self.storage_path.parent.mkdir(parents=True, exist_ok=True)
|
||||
data = [task.to_dict() for task in self.tasks.values()]
|
||||
with open(self.storage_path, 'w', encoding='utf-8') as f:
|
||||
json.dump(data, f, ensure_ascii=False, indent=2)
|
||||
|
||||
def register_action(self, action_name: str, handler: Callable):
|
||||
"""注册动作处理器"""
|
||||
self.action_handlers[action_name] = handler
|
||||
|
||||
def create_task(
|
||||
self,
|
||||
user_id: str,
|
||||
title: str,
|
||||
description: str,
|
||||
steps: List[Dict],
|
||||
metadata: Optional[Dict] = None
|
||||
) -> str:
|
||||
"""创建任务"""
|
||||
task_id = str(uuid.uuid4())
|
||||
|
||||
task_steps = [
|
||||
TaskStep(
|
||||
step_id=str(uuid.uuid4()),
|
||||
description=step['description'],
|
||||
action=step['action'],
|
||||
params=step.get('params', {})
|
||||
)
|
||||
for step in steps
|
||||
]
|
||||
|
||||
task = LongTask(
|
||||
task_id=task_id,
|
||||
user_id=user_id,
|
||||
title=title,
|
||||
description=description,
|
||||
steps=task_steps,
|
||||
metadata=metadata or {}
|
||||
)
|
||||
|
||||
self.tasks[task_id] = task
|
||||
self._save()
|
||||
|
||||
return task_id
|
||||
|
||||
async def execute_task(
|
||||
self,
|
||||
task_id: str,
|
||||
progress_callback: Optional[Callable[[str, float, str], None]] = None
|
||||
) -> bool:
|
||||
"""执行任务"""
|
||||
if task_id not in self.tasks:
|
||||
return False
|
||||
|
||||
task = self.tasks[task_id]
|
||||
|
||||
if task.status == TaskStatus.RUNNING:
|
||||
return False
|
||||
|
||||
task.status = TaskStatus.RUNNING
|
||||
task.started_at = datetime.now()
|
||||
self._save()
|
||||
|
||||
try:
|
||||
total_steps = len(task.steps)
|
||||
|
||||
for i, step in enumerate(task.steps):
|
||||
# 检查是否被取消
|
||||
if task.status == TaskStatus.CANCELLED:
|
||||
break
|
||||
|
||||
step.status = TaskStatus.RUNNING
|
||||
step.started_at = datetime.now()
|
||||
|
||||
# 执行步骤
|
||||
try:
|
||||
handler = self.action_handlers.get(step.action)
|
||||
if not handler:
|
||||
raise ValueError(f"未找到动作处理器: {step.action}")
|
||||
|
||||
result = await handler(**step.params)
|
||||
step.result = result
|
||||
step.status = TaskStatus.COMPLETED
|
||||
|
||||
except Exception as e:
|
||||
step.error = str(e)
|
||||
step.status = TaskStatus.FAILED
|
||||
task.status = TaskStatus.FAILED
|
||||
self._save()
|
||||
return False
|
||||
|
||||
finally:
|
||||
step.completed_at = datetime.now()
|
||||
|
||||
# 更新进度
|
||||
task.progress = (i + 1) / total_steps
|
||||
self._save()
|
||||
|
||||
if progress_callback:
|
||||
await progress_callback(
|
||||
task_id,
|
||||
task.progress,
|
||||
f"完成步骤 {i+1}/{total_steps}: {step.description}"
|
||||
)
|
||||
|
||||
task.status = TaskStatus.COMPLETED
|
||||
task.completed_at = datetime.now()
|
||||
task.progress = 1.0
|
||||
self._save()
|
||||
|
||||
return True
|
||||
|
||||
except Exception as e:
|
||||
task.status = TaskStatus.FAILED
|
||||
self._save()
|
||||
raise e
|
||||
|
||||
async def start_task(
|
||||
self,
|
||||
task_id: str,
|
||||
progress_callback: Optional[Callable[[str, float, str], None]] = None
|
||||
):
|
||||
"""启动任务（异步）"""
|
||||
if task_id in self.running_tasks:
|
||||
return
|
||||
|
||||
async def run():
|
||||
try:
|
||||
await self.execute_task(task_id, progress_callback)
|
||||
finally:
|
||||
if task_id in self.running_tasks:
|
||||
del self.running_tasks[task_id]
|
||||
|
||||
self.running_tasks[task_id] = asyncio.create_task(run())
|
||||
|
||||
def pause_task(self, task_id: str) -> bool:
|
||||
"""暂停任务"""
|
||||
if task_id not in self.tasks:
|
||||
return False
|
||||
|
||||
task = self.tasks[task_id]
|
||||
if task.status == TaskStatus.RUNNING:
|
||||
task.status = TaskStatus.PAUSED
|
||||
self._save()
|
||||
return True
|
||||
|
||||
return False
|
||||
|
||||
def cancel_task(self, task_id: str) -> bool:
|
||||
"""取消任务"""
|
||||
if task_id not in self.tasks:
|
||||
return False
|
||||
|
||||
task = self.tasks[task_id]
|
||||
task.status = TaskStatus.CANCELLED
|
||||
self._save()
|
||||
|
||||
# 取消正在运行的任务
|
||||
if task_id in self.running_tasks:
|
||||
self.running_tasks[task_id].cancel()
|
||||
|
||||
return True
|
||||
|
||||
def get_task(self, task_id: str) -> Optional[LongTask]:
|
||||
"""获取任务"""
|
||||
return self.tasks.get(task_id)
|
||||
|
||||
def get_user_tasks(self, user_id: str) -> List[LongTask]:
|
||||
"""获取用户的所有任务"""
|
||||
return [
|
||||
task for task in self.tasks.values()
|
||||
if task.user_id == user_id
|
||||
]
|
||||
|
||||
def get_task_status(self, task_id: str) -> Optional[Dict]:
|
||||
"""获取任务状态"""
|
||||
task = self.get_task(task_id)
|
||||
if not task:
|
||||
return None
|
||||
|
||||
completed_steps = sum(1 for step in task.steps if step.status == TaskStatus.COMPLETED)
|
||||
total_steps = len(task.steps)
|
||||
|
||||
return {
|
||||
'task_id': task.task_id,
|
||||
'title': task.title,
|
||||
'status': task.status.value,
|
||||
'progress': task.progress,
|
||||
'completed_steps': completed_steps,
|
||||
'total_steps': total_steps,
|
||||
'current_step': next(
|
||||
(step.description for step in task.steps if step.status == TaskStatus.RUNNING),
|
||||
None
|
||||
)
|
||||
}
|
||||
13
src/ai/vector_store/__init__.py
Normal file
13
src/ai/vector_store/__init__.py
Normal file
@@ -0,0 +1,13 @@
|
||||
"""
|
||||
向量存储模块
|
||||
"""
|
||||
from .base import VectorStore, VectorMemory
|
||||
from .chroma_store import ChromaVectorStore
|
||||
from .json_store import JSONVectorStore
|
||||
|
||||
__all__ = [
|
||||
'VectorStore',
|
||||
'VectorMemory',
|
||||
'ChromaVectorStore',
|
||||
'JSONVectorStore'
|
||||
]
|
||||
113
src/ai/vector_store/base.py
Normal file
113
src/ai/vector_store/base.py
Normal file
@@ -0,0 +1,113 @@
|
||||
"""
|
||||
向量数据库抽象层
|
||||
"""
|
||||
from abc import ABC, abstractmethod
|
||||
from typing import List, Dict, Optional, Tuple
|
||||
from dataclasses import dataclass
|
||||
from datetime import datetime
|
||||
|
||||
|
||||
@dataclass
|
||||
class VectorMemory:
|
||||
"""向量记忆项"""
|
||||
id: str
|
||||
user_id: str
|
||||
content: str
|
||||
embedding: List[float]
|
||||
importance: float
|
||||
timestamp: datetime
|
||||
metadata: Dict
|
||||
access_count: int = 0
|
||||
last_access: Optional[datetime] = None
|
||||
|
||||
def to_dict(self) -> Dict:
|
||||
"""转换为字典"""
|
||||
return {
|
||||
'id': self.id,
|
||||
'user_id': self.user_id,
|
||||
'content': self.content,
|
||||
'embedding': self.embedding,
|
||||
'importance': self.importance,
|
||||
'timestamp': self.timestamp.isoformat(),
|
||||
'metadata': self.metadata,
|
||||
'access_count': self.access_count,
|
||||
'last_access': self.last_access.isoformat() if self.last_access else None
|
||||
}
|
||||
|
||||
@classmethod
|
||||
def from_dict(cls, data: Dict) -> 'VectorMemory':
|
||||
"""从字典创建"""
|
||||
return cls(
|
||||
id=data['id'],
|
||||
user_id=data['user_id'],
|
||||
content=data['content'],
|
||||
embedding=data['embedding'],
|
||||
importance=data['importance'],
|
||||
timestamp=datetime.fromisoformat(data['timestamp']),
|
||||
metadata=data.get('metadata', {}),
|
||||
access_count=data.get('access_count', 0),
|
||||
last_access=datetime.fromisoformat(data['last_access']) if data.get('last_access') else None
|
||||
)
|
||||
|
||||
|
||||
class VectorStore(ABC):
|
||||
"""向量存储抽象基类"""
|
||||
|
||||
@abstractmethod
|
||||
async def add(
|
||||
self,
|
||||
id: str,
|
||||
user_id: str,
|
||||
content: str,
|
||||
embedding: List[float],
|
||||
importance: float,
|
||||
metadata: Optional[Dict] = None
|
||||
) -> bool:
|
||||
"""添加记忆"""
|
||||
pass
|
||||
|
||||
@abstractmethod
|
||||
async def search(
|
||||
self,
|
||||
user_id: str,
|
||||
query_embedding: List[float],
|
||||
limit: int = 5,
|
||||
min_importance: float = 0.3
|
||||
) -> List[VectorMemory]:
|
||||
"""搜索相似记忆"""
|
||||
pass
|
||||
|
||||
@abstractmethod
|
||||
async def get_by_importance(
|
||||
self,
|
||||
user_id: str,
|
||||
limit: int = 5,
|
||||
min_importance: float = 0.3
|
||||
) -> List[VectorMemory]:
|
||||
"""按重要性获取记忆"""
|
||||
pass
|
||||
|
||||
@abstractmethod
|
||||
async def update_access(self, memory_id: str) -> bool:
|
||||
"""更新访问记录"""
|
||||
pass
|
||||
|
||||
@abstractmethod
|
||||
async def delete(self, memory_id: str) -> bool:
|
||||
"""删除记忆"""
|
||||
pass
|
||||
|
||||
@abstractmethod
|
||||
async def get_all(self, user_id: str) -> List[VectorMemory]:
|
||||
"""获取用户所有记忆"""
|
||||
pass
|
||||
|
||||
@abstractmethod
|
||||
async def clear_user(self, user_id: str) -> bool:
|
||||
"""清除用户所有记忆"""
|
||||
pass
|
||||
|
||||
@abstractmethod
|
||||
async def close(self):
|
||||
"""关闭连接"""
|
||||
pass
|
||||
336
src/ai/vector_store/chroma_store.py
Normal file
336
src/ai/vector_store/chroma_store.py
Normal file
@@ -0,0 +1,336 @@
|
||||
"""
|
||||
Chroma向量数据库实现
|
||||
"""
|
||||
import uuid
|
||||
from typing import List, Dict, Optional
|
||||
from pathlib import Path
|
||||
from datetime import datetime
|
||||
import chromadb
|
||||
from chromadb.config import Settings
|
||||
from .base import VectorStore, VectorMemory
|
||||
from src.utils.logger import setup_logger
|
||||
|
||||
logger = setup_logger('ChromaStore')
|
||||
|
||||
|
||||
class ChromaVectorStore(VectorStore):
|
||||
"""Chroma向量数据库实现"""
|
||||
|
||||
def __init__(self, persist_directory: Path):
|
||||
"""初始化Chroma客户端"""
|
||||
self.persist_directory = persist_directory
|
||||
self.persist_directory.mkdir(parents=True, exist_ok=True)
|
||||
|
||||
# 创建Chroma客户端
|
||||
self.client = chromadb.PersistentClient(
|
||||
path=str(persist_directory),
|
||||
settings=Settings(
|
||||
anonymized_telemetry=False,
|
||||
allow_reset=True
|
||||
)
|
||||
)
|
||||
|
||||
# 获取或创建集合
|
||||
self.collection = self.client.get_or_create_collection(
|
||||
name="long_term_memory",
|
||||
metadata={"description": "Long-term memory storage"}
|
||||
)
|
||||
|
||||
logger.info(f"✅ Chroma向量数据库初始化: {persist_directory}")
|
||||
|
||||
@staticmethod
|
||||
def _to_list(value):
|
||||
"""将可能的 numpy 数组等序列转为 Python list。"""
|
||||
if value is None:
|
||||
return []
|
||||
if hasattr(value, "tolist"):
|
||||
return value.tolist()
|
||||
if isinstance(value, list):
|
||||
return value
|
||||
return list(value)
|
||||
|
||||
async def add(
|
||||
self,
|
||||
id: str,
|
||||
user_id: str,
|
||||
content: str,
|
||||
embedding: List[float],
|
||||
importance: float,
|
||||
metadata: Optional[Dict] = None
|
||||
) -> bool:
|
||||
"""添加记忆"""
|
||||
try:
|
||||
if not embedding:
|
||||
logger.warning(f"跳过写入长期记忆（embedding为空）: {id}")
|
||||
return False
|
||||
|
||||
now = datetime.now()
|
||||
|
||||
# 准备元数据
|
||||
meta = {
|
||||
'user_id': user_id,
|
||||
'importance': importance,
|
||||
'timestamp': now.isoformat(),
|
||||
'access_count': 0,
|
||||
'last_access': '',
|
||||
}
|
||||
|
||||
if metadata:
|
||||
# 只保存可序列化的元数据
|
||||
for key, value in metadata.items():
|
||||
if isinstance(value, (str, int, float, bool)):
|
||||
meta[f'meta_{key}'] = value
|
||||
|
||||
# 添加到Chroma
|
||||
self.collection.add(
|
||||
ids=[id],
|
||||
embeddings=[embedding],
|
||||
documents=[content],
|
||||
metadatas=[meta]
|
||||
)
|
||||
|
||||
logger.debug(f"添加记忆: {id} (用户: {user_id})")
|
||||
return True
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"添加记忆失败: {e}")
|
||||
return False
|
||||
|
||||
async def search(
|
||||
self,
|
||||
user_id: str,
|
||||
query_embedding: List[float],
|
||||
limit: int = 5,
|
||||
min_importance: float = 0.3
|
||||
) -> List[VectorMemory]:
|
||||
"""搜索相似记忆"""
|
||||
try:
|
||||
# 查询Chroma
|
||||
results = self.collection.query(
|
||||
query_embeddings=[query_embedding],
|
||||
n_results=limit * 2, # 多查询一些，用于过滤
|
||||
where={"user_id": user_id}
|
||||
)
|
||||
|
||||
ids = self._to_list(results.get('ids'))
|
||||
if len(ids) == 0:
|
||||
return []
|
||||
first_row_ids = self._to_list(ids[0])
|
||||
if len(first_row_ids) == 0:
|
||||
return []
|
||||
|
||||
metadatas = self._to_list(results.get('metadatas'))
|
||||
documents = self._to_list(results.get('documents'))
|
||||
embeddings = self._to_list(results.get('embeddings'))
|
||||
first_row_metas = self._to_list(metadatas[0]) if len(metadatas) > 0 else []
|
||||
first_row_docs = self._to_list(documents[0]) if len(documents) > 0 else []
|
||||
first_row_embeddings = self._to_list(embeddings[0]) if len(embeddings) > 0 else []
|
||||
|
||||
# 转换结果
|
||||
memories = []
|
||||
for i in range(len(first_row_ids)):
|
||||
if i >= len(first_row_metas) or i >= len(first_row_docs):
|
||||
continue
|
||||
|
||||
meta = first_row_metas[i]
|
||||
|
||||
# 过滤重要性
|
||||
if meta['importance'] < min_importance:
|
||||
continue
|
||||
|
||||
# 提取自定义元数据
|
||||
custom_meta = {}
|
||||
for key, value in meta.items():
|
||||
if key.startswith('meta_'):
|
||||
custom_meta[key[5:]] = value
|
||||
|
||||
memory = VectorMemory(
|
||||
id=first_row_ids[i],
|
||||
user_id=meta['user_id'],
|
||||
content=first_row_docs[i],
|
||||
embedding=self._to_list(first_row_embeddings[i]) if i < len(first_row_embeddings) else [],
|
||||
importance=meta['importance'],
|
||||
timestamp=datetime.fromisoformat(meta['timestamp']),
|
||||
metadata=custom_meta,
|
||||
access_count=meta.get('access_count', 0),
|
||||
last_access=datetime.fromisoformat(meta['last_access']) if meta.get('last_access') else None
|
||||
)
|
||||
memories.append(memory)
|
||||
|
||||
if len(memories) >= limit:
|
||||
break
|
||||
|
||||
return memories
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"搜索记忆失败: {e}")
|
||||
return []
|
||||
|
||||
async def get_by_importance(
|
||||
self,
|
||||
user_id: str,
|
||||
limit: int = 5,
|
||||
min_importance: float = 0.3
|
||||
) -> List[VectorMemory]:
|
||||
"""按重要性获取记忆"""
|
||||
try:
|
||||
# 获取所有用户记忆
|
||||
results = self.collection.get(
|
||||
where={"user_id": user_id},
|
||||
include=['embeddings', 'documents', 'metadatas']
|
||||
)
|
||||
|
||||
ids = self._to_list(results.get('ids'))
|
||||
if len(ids) == 0:
|
||||
return []
|
||||
|
||||
metadatas = self._to_list(results.get('metadatas'))
|
||||
documents = self._to_list(results.get('documents'))
|
||||
embeddings = self._to_list(results.get('embeddings'))
|
||||
|
||||
# 转换并排序
|
||||
memories = []
|
||||
for i in range(len(ids)):
|
||||
if i >= len(metadatas) or i >= len(documents):
|
||||
continue
|
||||
meta = metadatas[i]
|
||||
|
||||
if meta['importance'] < min_importance:
|
||||
continue
|
||||
|
||||
custom_meta = {}
|
||||
for key, value in meta.items():
|
||||
if key.startswith('meta_'):
|
||||
custom_meta[key[5:]] = value
|
||||
|
||||
memory = VectorMemory(
|
||||
id=ids[i],
|
||||
user_id=meta['user_id'],
|
||||
content=documents[i],
|
||||
embedding=self._to_list(embeddings[i]) if i < len(embeddings) else [],
|
||||
importance=meta['importance'],
|
||||
timestamp=datetime.fromisoformat(meta['timestamp']),
|
||||
metadata=custom_meta,
|
||||
access_count=meta.get('access_count', 0),
|
||||
last_access=datetime.fromisoformat(meta['last_access']) if meta.get('last_access') else None
|
||||
)
|
||||
memories.append(memory)
|
||||
|
||||
# 按重要性和时间排序
|
||||
memories.sort(key=lambda m: (m.importance, m.timestamp), reverse=True)
|
||||
return memories[:limit]
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"获取记忆失败: {e}")
|
||||
return []
|
||||
|
||||
async def update_access(self, memory_id: str) -> bool:
|
||||
"""更新访问记录"""
|
||||
try:
|
||||
# 获取当前记忆
|
||||
result = self.collection.get(
|
||||
ids=[memory_id],
|
||||
include=['metadatas']
|
||||
)
|
||||
|
||||
ids = self._to_list(result.get('ids'))
|
||||
if len(ids) == 0:
|
||||
return False
|
||||
|
||||
metadatas = self._to_list(result.get('metadatas'))
|
||||
if len(metadatas) == 0:
|
||||
return False
|
||||
meta = metadatas[0]
|
||||
meta['access_count'] = meta.get('access_count', 0) + 1
|
||||
meta['last_access'] = datetime.now().isoformat()
|
||||
|
||||
# 更新元数据
|
||||
self.collection.update(
|
||||
ids=[memory_id],
|
||||
metadatas=[meta]
|
||||
)
|
||||
|
||||
return True
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"更新访问记录失败: {e}")
|
||||
return False
|
||||
|
||||
async def delete(self, memory_id: str) -> bool:
|
||||
"""删除记忆"""
|
||||
try:
|
||||
self.collection.delete(ids=[memory_id])
|
||||
return True
|
||||
except Exception as e:
|
||||
logger.error(f"删除记忆失败: {e}")
|
||||
return False
|
||||
|
||||
async def get_all(self, user_id: str) -> List[VectorMemory]:
|
||||
"""获取用户所有记忆"""
|
||||
try:
|
||||
results = self.collection.get(
|
||||
where={"user_id": user_id},
|
||||
include=['embeddings', 'documents', 'metadatas']
|
||||
)
|
||||
|
||||
ids = self._to_list(results.get('ids'))
|
||||
if len(ids) == 0:
|
||||
return []
|
||||
|
||||
metadatas = self._to_list(results.get('metadatas'))
|
||||
documents = self._to_list(results.get('documents'))
|
||||
embeddings = self._to_list(results.get('embeddings'))
|
||||
|
||||
memories = []
|
||||
for i in range(len(ids)):
|
||||
if i >= len(metadatas) or i >= len(documents):
|
||||
continue
|
||||
meta = metadatas[i]
|
||||
|
||||
custom_meta = {}
|
||||
for key, value in meta.items():
|
||||
if key.startswith('meta_'):
|
||||
custom_meta[key[5:]] = value
|
||||
|
||||
memory = VectorMemory(
|
||||
id=ids[i],
|
||||
user_id=meta['user_id'],
|
||||
content=documents[i],
|
||||
embedding=self._to_list(embeddings[i]) if i < len(embeddings) else [],
|
||||
importance=meta['importance'],
|
||||
timestamp=datetime.fromisoformat(meta['timestamp']),
|
||||
metadata=custom_meta,
|
||||
access_count=meta.get('access_count', 0),
|
||||
last_access=datetime.fromisoformat(meta['last_access']) if meta.get('last_access') else None
|
||||
)
|
||||
memories.append(memory)
|
||||
|
||||
return memories
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"获取所有记忆失败: {e}")
|
||||
return []
|
||||
|
||||
async def clear_user(self, user_id: str) -> bool:
|
||||
"""清除用户所有记忆"""
|
||||
try:
|
||||
# 获取用户所有记忆ID
|
||||
results = self.collection.get(
|
||||
where={"user_id": user_id}
|
||||
)
|
||||
|
||||
ids = self._to_list(results.get('ids'))
|
||||
if len(ids) > 0:
|
||||
self.collection.delete(ids=ids)
|
||||
|
||||
logger.info(f"清除用户记忆: {user_id}")
|
||||
return True
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"清除用户记忆失败: {e}")
|
||||
return False
|
||||
|
||||
async def close(self):
|
||||
"""关闭连接"""
|
||||
# Chroma会自动持久化，无需特殊关闭操作
|
||||
logger.info("Chroma向量数据库已关闭")
|
||||
198
src/ai/vector_store/json_store.py
Normal file
198
src/ai/vector_store/json_store.py
Normal file
@@ -0,0 +1,198 @@
|
||||
"""
|
||||
JSON文件存储实现（向后兼容）
|
||||
"""
|
||||
import json
|
||||
import uuid
|
||||
from typing import List, Dict, Optional
|
||||
from pathlib import Path
|
||||
from datetime import datetime
|
||||
import numpy as np
|
||||
from .base import VectorStore, VectorMemory
|
||||
from src.utils.logger import setup_logger
|
||||
|
||||
logger = setup_logger('JSONStore')
|
||||
|
||||
|
||||
class JSONVectorStore(VectorStore):
|
||||
"""JSON文件存储实现（向后兼容旧版本）"""
|
||||
|
||||
def __init__(self, storage_path: Path):
|
||||
"""初始化JSON存储"""
|
||||
self.storage_path = storage_path
|
||||
self.memories: Dict[str, List[VectorMemory]] = {} # user_id -> List[VectorMemory]
|
||||
self._load()
|
||||
logger.info(f"✅ JSON存储初始化: {storage_path}")
|
||||
|
||||
def _load(self):
|
||||
"""加载记忆"""
|
||||
if self.storage_path.exists():
|
||||
try:
|
||||
with open(self.storage_path, 'r', encoding='utf-8') as f:
|
||||
data = json.load(f)
|
||||
|
||||
for user_id, items in data.items():
|
||||
self.memories[user_id] = []
|
||||
for item in items:
|
||||
# 兼容旧格式
|
||||
if 'id' not in item:
|
||||
item['id'] = str(uuid.uuid4())
|
||||
|
||||
memory = VectorMemory.from_dict(item)
|
||||
self.memories[user_id].append(memory)
|
||||
|
||||
logger.info(f"加载了 {sum(len(v) for v in self.memories.values())} 条记忆")
|
||||
except Exception as e:
|
||||
logger.error(f"加载记忆失败: {e}")
|
||||
self.memories = {}
|
||||
|
||||
def _save(self):
|
||||
"""保存记忆"""
|
||||
try:
|
||||
self.storage_path.parent.mkdir(parents=True, exist_ok=True)
|
||||
data = {
|
||||
user_id: [memory.to_dict() for memory in memories]
|
||||
for user_id, memories in self.memories.items()
|
||||
}
|
||||
|
||||
with open(self.storage_path, 'w', encoding='utf-8') as f:
|
||||
json.dump(data, f, ensure_ascii=False, indent=2)
|
||||
except Exception as e:
|
||||
logger.error(f"保存记忆失败: {e}")
|
||||
|
||||
async def add(
|
||||
self,
|
||||
id: str,
|
||||
user_id: str,
|
||||
content: str,
|
||||
embedding: List[float],
|
||||
importance: float,
|
||||
metadata: Optional[Dict] = None
|
||||
) -> bool:
|
||||
"""添加记忆"""
|
||||
try:
|
||||
memory = VectorMemory(
|
||||
id=id,
|
||||
user_id=user_id,
|
||||
content=content,
|
||||
embedding=embedding,
|
||||
importance=importance,
|
||||
timestamp=datetime.now(),
|
||||
metadata=metadata or {},
|
||||
access_count=0,
|
||||
last_access=None
|
||||
)
|
||||
|
||||
if user_id not in self.memories:
|
||||
self.memories[user_id] = []
|
||||
|
||||
self.memories[user_id].append(memory)
|
||||
self._save()
|
||||
|
||||
logger.debug(f"添加记忆: {id} (用户: {user_id})")
|
||||
return True
|
||||
except Exception as e:
|
||||
logger.error(f"添加记忆失败: {e}")
|
||||
return False
|
||||
|
||||
async def search(
|
||||
self,
|
||||
user_id: str,
|
||||
query_embedding: List[float],
|
||||
limit: int = 5,
|
||||
min_importance: float = 0.3
|
||||
) -> List[VectorMemory]:
|
||||
"""搜索相似记忆"""
|
||||
if user_id not in self.memories:
|
||||
return []
|
||||
|
||||
memories = self.memories[user_id]
|
||||
|
||||
# 过滤重要性
|
||||
memories = [m for m in memories if m.importance >= min_importance]
|
||||
|
||||
if not memories:
|
||||
return []
|
||||
|
||||
# 使用向量相似度排序
|
||||
scored_memories = []
|
||||
for memory in memories:
|
||||
if memory.embedding:
|
||||
similarity = self._cosine_similarity(query_embedding, memory.embedding)
|
||||
scored_memories.append((similarity, memory))
|
||||
|
||||
if not scored_memories:
|
||||
# 如果没有嵌入向量，按重要性排序
|
||||
return await self.get_by_importance(user_id, limit, min_importance)
|
||||
|
||||
scored_memories.sort(reverse=True, key=lambda x: x[0])
|
||||
return [m for _, m in scored_memories[:limit]]
|
||||
|
||||
async def get_by_importance(
|
||||
self,
|
||||
user_id: str,
|
||||
limit: int = 5,
|
||||
min_importance: float = 0.3
|
||||
) -> List[VectorMemory]:
|
||||
"""按重要性获取记忆"""
|
||||
if user_id not in self.memories:
|
||||
return []
|
||||
|
||||
memories = [m for m in self.memories[user_id] if m.importance >= min_importance]
|
||||
memories.sort(key=lambda m: (m.importance, m.timestamp), reverse=True)
|
||||
return memories[:limit]
|
||||
|
||||
def _cosine_similarity(self, vec1: List[float], vec2: List[float]) -> float:
|
||||
"""计算余弦相似度"""
|
||||
vec1 = np.array(vec1)
|
||||
vec2 = np.array(vec2)
|
||||
return np.dot(vec1, vec2) / (np.linalg.norm(vec1) * np.linalg.norm(vec2))
|
||||
|
||||
async def update_access(self, memory_id: str) -> bool:
|
||||
"""更新访问记录"""
|
||||
try:
|
||||
for memories in self.memories.values():
|
||||
for memory in memories:
|
||||
if memory.id == memory_id:
|
||||
memory.access_count += 1
|
||||
memory.last_access = datetime.now()
|
||||
self._save()
|
||||
return True
|
||||
return False
|
||||
except Exception as e:
|
||||
logger.error(f"更新访问记录失败: {e}")
|
||||
return False
|
||||
|
||||
async def delete(self, memory_id: str) -> bool:
|
||||
"""删除记忆"""
|
||||
try:
|
||||
for user_id, memories in self.memories.items():
|
||||
for i, memory in enumerate(memories):
|
||||
if memory.id == memory_id:
|
||||
del self.memories[user_id][i]
|
||||
self._save()
|
||||
return True
|
||||
return False
|
||||
except Exception as e:
|
||||
logger.error(f"删除记忆失败: {e}")
|
||||
return False
|
||||
|
||||
async def get_all(self, user_id: str) -> List[VectorMemory]:
|
||||
"""获取用户所有记忆"""
|
||||
return self.memories.get(user_id, [])
|
||||
|
||||
async def clear_user(self, user_id: str) -> bool:
|
||||
"""清除用户所有记忆"""
|
||||
try:
|
||||
if user_id in self.memories:
|
||||
del self.memories[user_id]
|
||||
self._save()
|
||||
logger.info(f"清除用户记忆: {user_id}")
|
||||
return True
|
||||
except Exception as e:
|
||||
logger.error(f"清除用户记忆失败: {e}")
|
||||
return False
|
||||
|
||||
async def close(self):
|
||||
"""关闭连接"""
|
||||
self._save()
|
||||
logger.info("JSON存储已关闭")
|
||||
Reference in New Issue
Block a user