ワンクリックで
context7
Fetch up-to-date library documentation from Context7 API. Use when user asks "how do I use this library", "show me docs for", "library documentation", "API reference for", or needs current documentation for any code library.
Fetch up-to-date library documentation from Context7 API. Use when user asks "how do I use this library", "show me docs for", "library documentation", "API reference for", or needs current documentation for any code library.
File-based inter-agent messaging. Check inbox, send bugs/requests to other projects, acknowledge resolved issues. Use for cross-project agent communication.
Search arXiv for academic papers. Use when user says "find papers on", "search arxiv", "look up research", "academic papers about", "what papers exist on", "use your arxiv tool", or asks about scientific/ML research topics.
Step back and critically reassess project state. Use when asked to "assess", "step back", "fresh eyes", "check alignment", "sanity check", "health check", or "evaluate what's working". Offer to run after major changes (don't auto-run).
Free web and local search via Brave Search API. Use when user says "brave search", "search with brave", "brave web search", "brave local search", "local search", "find businesses near", or "near me".
Submit code review requests to multiple AI providers (GitHub Copilot, Anthropic Claude, OpenAI Codex, Google Gemini) and get patches back. Use when user says "code review", "review this code", "get a patch for", or needs AI-generated unified diffs for code fixes.
Distill PDF, URL, or text into Q&A pairs stored in memory. Use --context for domain-focused extraction.
| name | context7 |
| description | Fetch up-to-date library documentation from Context7 API. Use when user asks "how do I use this library", "show me docs for", "library documentation", "API reference for", or needs current documentation for any code library. |
| allowed-tools | Bash, Read |
| triggers | ["library documentation","show me docs for","API reference","how to use this library","latest docs for","context7 lookup"] |
| metadata | {"short-description":"Library documentation lookup via Context7"} |
Fetch up-to-date library documentation from Context7 API for ANY code library.
CONTEXT7_API_KEY environment variable set (check .env)# Step 1: Find the library ID
python .agents/skills/context7/context7.py search <library-name> "<your-query>"
# Step 2: Get documentation context
python .agents/skills/context7/context7.py context <library-id> "<your-query>" --tokens 5000
| Command | Description | Example |
|---|---|---|
search | Rank repositories by relevance to your query | python .agents/skills/context7/context7.py search arangodb "bm25 search" |
context | Download reranked doc chunks for a specific library ID | python .agents/skills/context7/context7.py context /arangodb/arangodb "vector search" |
find | Convenience multi-library search across common stacks | python .agents/skills/context7/context7.py find "binary search" |
These are the only supported subcommands; invoking the script without one will show Typer’s usage error. The table mirrors the exact signatures implemented in context7.py, so copy/paste examples will work as-is.
Find libraries by name with LLM-powered ranking. Works with ANY library on GitHub:
# Search for any library - just change the libraryName
curl -s -X GET "https://context7.com/api/v2/libs/search?libraryName=<YOUR-LIBRARY>&query=<your-query>" \
-H "Authorization: Bearer $CONTEXT7_API_KEY" | jq '.results[:3]'
# Examples for different libraries:
curl -s "https://context7.com/api/v2/libs/search?libraryName=pandas&query=dataframe+merge" ...
curl -s "https://context7.com/api/v2/libs/search?libraryName=tensorflow&query=keras+model" ...
curl -s "https://context7.com/api/v2/libs/search?libraryName=django&query=orm+query" ...
Response includes library IDs like /owner/repo that you use in the context endpoint.
Retrieve LLM-reranked documentation snippets for a query:
# Get ArangoDB BM25 documentation
curl -s -X GET "https://context7.com/api/v2/context?libraryId=/arangodb/arangodb&query=bm25+search+arangosearch&tokens=5000" \
-H "Authorization: Bearer $CONTEXT7_API_KEY"
# Get Lean4 tactic documentation
curl -s -X GET "https://context7.com/api/v2/context?libraryId=/leanprover/lean4&query=simp+tactic&tokens=3000" \
-H "Authorization: Bearer $CONTEXT7_API_KEY"
# Get sentence-transformers embedding docs
curl -s -X GET "https://context7.com/api/v2/context?libraryId=/UKPLab/sentence-transformers&query=encode+embeddings+cosine&tokens=3000" \
-H "Authorization: Bearer $CONTEXT7_API_KEY"
Parameters:
libraryId: Library ID from search (e.g., /arangodb/arangodb)query: Natural language querytokens: Max tokens to return (default ~5000)Use python context7.py search <name> "<query>" to find ANY library's ID.
| Library | Library ID |
|---|---|
| ArangoDB | /arangodb/arangodb |
| Lean 4 | /leanprover/lean4 |
| sentence-transformers | /UKPLab/sentence-transformers |
| PyTorch | /pytorch/pytorch |
| TensorFlow | /tensorflow/tensorflow |
| Pandas | /pandas-dev/pandas |
| NumPy | /numpy/numpy |
| Django | /django/django |
| Flask | /pallets/flask |
| FastAPI | /tiangolo/fastapi |
| Next.js | /vercel/next.js |
| React | /facebook/react |
| Vue.js | /vuejs/vue |
| Svelte | /sveltejs/svelte |
| Express | /expressjs/express |
| Rust std | /rust-lang/rust |
| Go std | /golang/go |
CONTEXT7_API_KEY=$(grep CONTEXT7_API_KEY .env | cut -d= -f2) \
curl -s "https://context7.com/api/v2/context?libraryId=/arangodb/arangodb&query=cosine+similarity+vector+search&tokens=3000" \
-H "Authorization: Bearer $CONTEXT7_API_KEY"
CONTEXT7_API_KEY=$(grep CONTEXT7_API_KEY .env | cut -d= -f2) \
curl -s "https://context7.com/api/v2/context?libraryId=/leanprover/lean4&query=omega+tactic+natural+numbers&tokens=3000" \
-H "Authorization: Bearer $CONTEXT7_API_KEY"
The CLI surface is the preferred interface. If you embed it elsewhere, import the Typer app or the helper functions directly:
from .context7 import app, _search_libs, _get_context
result = _search_libs("arangodb", "bm25 search")
docs = _get_context("/arangodb/arangodb", "bm25 arangosearch scoring")
This matches the shipped code (context7.py).
.agents/skills/dotenv_helper.py loads .env automatically so CONTEXT7_API_KEY is present even when skills run via uvx..agents/skills/json_utils.py can be imported to repair JSON before forwarding it to downstream tooling if you extend the skill; the built-in CLI already prints valid JSON.