knowledge-rag
Operate on BiSheng knowledge libraries, knowledge spaces, document ingestion, parser providers, RAG pipeline, Milvus/Elasticsearch recall, and knowledge workers.
- License 许可
- Apache-2.0
- 最近更新
- 2026-10-08
- Stars
- ★ 11
- Forks
- 3
01 前言 Front Matter
- name
- knowledge-rag
- license
- Apache 2.0
- metadata
- { "disco-role": "operating" }
- description
- Operate on BiSheng knowledge libraries, knowledge spaces, document ingestion, parser providers, RAG pipeline, Milvus/Elasticsearch recall, and knowledge workers.
- disable-model-invocation
- true
02 资源文件 2
- 参考文档references/troubleshooting.md3.2 KB
- 参考文档references/workflows.md3.3 KB
03 SKILL.md 正文
knowledge-rag
Use this sub-skill when a task touches BiSheng knowledge libraries, knowledge spaces, QA libraries, file upload/parse states, document parser providers, the Load → Transform → Ingest pipeline, Milvus + Elasticsearch dual recall, MinIO file artifacts, or knowledge_celery worker behavior.
Start here
Run bundled helper commands from this sub-skill directory, or adjust the script path to this directory after import.
- Inspect config-shaped text before debugging parser/storage problems:
python scripts/check_knowledge_config.py --config <bisheng-checkout>/docker/bisheng/config/config.yaml - Read references/workflows.md for ingestion, parser, retrieval, and worker workflows.
- Read references/troubleshooting.md for parser, object storage, vector index, queue, and permission symptoms.
Owned responsibilities
- Knowledge API/domain services under
src/backend/bisheng/knowledge/. - RAG pipeline files under
knowledge/rag/, especially file pipelines, loader/provider selection, transformers, Milvus and Elasticsearch factories. bisheng_langchain/rag/retrievers, RAG pipeline, scoring, and rerank integration when used by BiSheng knowledge workflows.- Knowledge Celery workers under
src/backend/bisheng/worker/knowledge/. - Tests under
src/backend/test/knowledge/,test/knowledge/rag/, and safe config/service tests. - Frontend knowledge-space or knowledge-library callers only at routing depth; UI rendering details belong to
frontend-apps.
Route sibling areas instead of duplicating them
- Use
identity-permissions-tenancyfor space visibility, OpenFGA/ReBAC, approval, tenant filtering, and cursor permission scans. - Use
workflow-enginefor workflow RAG node execution outside the knowledge pipeline itself. - Use
backend-corefor generic FastAPI router/envelope/service conventions. - Use
frontend-appsfor Platform/Client knowledge pages and request wrappers. - Use
deployment-maintenancefor Docker storage service startup and production config rollout.
Non-negotiables
- Treat Load, Transform, and Ingest as separate stages when debugging: a loader success does not prove vector writes or preview cache.
- Preserve Milvus dense and Elasticsearch sparse/BM25 behavior when changing retrieval.
- Do not run external parser services, OCR downloads, or large document parsing as a smoke check unless the environment is approved.
- Use small fixtures for tests; do not require production MinIO/Milvus/ES for unit-level parser or service logic.
- Keep parser-provider credentials, endpoints, and object-storage secrets out of skill output and committed test fixtures.
04 同仓库更多 Skills
obsidian-clipper-template-creator
Guide for creating templates for the Obsidian Web Clipper. Use when you want to create a new clipping template, understand available variables, or format clipped content.
obsidian-clipper-template-creator
Guide for creating templates for the Obsidian Web Clipper. Use when you want to create a new clipping template, understand available variables, or format clipped content.
python-api
Programmatic Object-Detection-Metrics API guidance for BoundingBox, BoundingBoxes, Evaluator, enums, coordinate conversion, AP/mAP semantics, and source-style import diagnostics.
file-evaluation
Folder-based PASCAL VOC AP/mAP evaluation from ground-truth and detection text files.
object-detection-metrics
Guides legacy Object-Detection-Metrics PASCAL VOC AP/mAP evaluation from detection text folders and direct Python API classes.
nx-workspace-patterns
Configure and optimize Nx monorepo workspaces. Use when setting up Nx, configuring project boundaries, optimizing build caching, or implementing affected commands.