rag-and-tools
Use MetaGPT RAG pipelines, retrievers, document stores, search/browser/data tools, and tool registry diagnostics.
- License 许可
- MIT
- 最近更新
- 2026-10-08
- Stars
- ★ 11
- Forks
- 3
01 前言 Front Matter
- name
- rag-and-tools
- license
- MIT
- metadata
- { "disco-role": "operating" }
- description
- Use MetaGPT RAG pipelines, retrievers, document stores, search/browser/data tools, and tool registry diagnostics.
- disable-model-invocation
- true
02 资源文件 4
- 参考文档references/api-reference.md15.9 KB
- 参考文档references/dependencies.md12.3 KB
- 参考文档references/troubleshooting.md13.2 KB
- 参考文档references/workflows.md15.3 KB
03 SKILL.md 正文
MetaGPT RAG and Tools
Use this sub-skill when a request involves MetaGPT retrieval-augmented generation, vector/document stores, retriever/ranker selection, search or browser engines, data/editor/tool libraries, tool registration, tool recommendation, or safe optional-dependency diagnosis.
Route Here For
- Building a MetaGPT RAG pipeline from documents or objects with
metagpt.rag.engines.SimpleEngine. - Choosing or debugging retrievers, rankers, vector stores, persisted indexes, and document-store wrappers.
- Using web/search/browser/editor/data tools from
metagpt.toolsandmetagpt.tools.libs. - Registering custom tools with
ToolRegistry, selecting registered tools by name/tag/path, or using tool recommenders. - Diagnosing missing optional packages for
[rag], search providers, browser engines, vector stores, and tool recommendation.
Route Elsewhere
- Data Interpreter orchestration, plans, and DI execution loops: use the
data-interpretersub-skill; return here only for tool prerequisites or tool-specific failures. - Core software-company CLI,
Role/Actionbasics, and multi-role project workflows: usesoftware-company. - Extension optimizers, Android/browser environments, and environment integrations outside RAG/tools: use
extensions-and-environments. - Maintainer serialization, memory, experience pools, and internal persistence APIs: use
maintainer-apis. - Installation/config basics before optional services or API keys exist: use the root MetaGPT installation/config guidance, then return here.
Reference Map
references/workflows.md: RAG construction, indexing, querying, retriever/ranker choices, search/browser usage, and tool registration recipes.references/api-reference.md: key modules, classes, config schemas, factories, and registry/recommender concepts.references/dependencies.md: optional extras, package groups, service/key/browser prerequisites, and missing-dependency symptoms.references/troubleshooting.md: vector-store, embedding/LLM, search, browser, data schema, and tool-safety failure handling.scripts/rag_import_check.py: safe import/spec diagnostic helper for selected RAG/tool modules and optional dependencies.
Quick Start
- Read
references/dependencies.mdfirst if base MetaGPT is installed withoutmetagpt[rag], search extras, browser extras, or vector-store clients. - Use
scripts/rag_import_check.py --group ragor--group toolsto identify missing optional packages before writing code that imports deep RAG/tool modules. - Use
references/workflows.mdfor minimal recipes; avoid running network search, browser automation, external vector DBs, or long LLM/reranker calls unless the user confirms prerequisites and safety.
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.