Context7 Documentation Retrieval

This skill fetches fresh documentation snippets from Context7. Use it instead of relying on possibly outdated training data.

When to use

•The user asks for docs, API signatures, configuration options, or examples for a library/framework.
•You need to verify correct usage or behavior in a recent version.
•The question is specific to a feature, function, or component.

Inputs to gather

•Library name (required): e.g., react, nextjs, fastapi, axios
•Topic (required): e.g., "useState", "app router", "dependency injection"
•Version (optional): e.g., "v5", "React 19", "Next 14" (append to query)
•Output preference (optional): examples, migration notes, configuration, etc.

If the user does not provide a topic, ask for the exact feature, API, or task.

Workflow

1) Find the library ID

Search for the best matching library. Use a narrow topic to improve ranking.

bash

curl -sS "https://context7.com/api/v2/libs/search?libraryName=react&query=hooks" | jq '.results[0]'

If jq is unavailable, use Python:

bash

python3 - <<'PY'
import json,sys
import urllib.request
url = "https://context7.com/api/v2/libs/search?libraryName=react&query=hooks"
data = json.load(urllib.request.urlopen(url))
print(json.dumps(data.get("results", [None])[0], indent=2))
PY

Disambiguation tips

•Prefer results with titles that match the official docs.
•Use description and totalSnippets to confirm relevance.
•If unclear, inspect .results[0..3] and pick the closest match.
•If still ambiguous, present top options and ask the user to choose.

2) Fetch documentation

Use the chosen libraryId and a focused query. Default to type=txt for readability.

bash

curl -sS "https://context7.com/api/v2/context?libraryId=/websites/react_dev_reference&query=useState&type=txt"

If the user mentions a version, add it to the query:

bash

curl -sS "https://context7.com/api/v2/context?libraryId=/vercel/next.js&query=app+router+next+14&type=txt"

If you need structured output, use type=json and extract snippets.

3) Answer with evidence

Provide a concise summary and quote the relevant lines from the snippet. Do not invent APIs.

Recommended output structure:

•2-5 bullet summary of what the docs say
•short quoted snippet(s) as evidence
•minimal code example if the user asked for one
•note any uncertainty or missing details

4) If results are empty or off-topic

•Broaden or rephrase the query (e.g., add the component name).
•Try alternate library names (reactjs vs react, next vs nextjs).
•Check additional search results and pick a better libraryId.
•Ask a focused follow-up question only if necessary.

Examples

React hooks documentation

bash

curl -sS "https://context7.com/api/v2/libs/search?libraryName=react&query=hooks" | jq '.results[0].id'
curl -sS "https://context7.com/api/v2/context?libraryId=/websites/react_dev_reference&query=useState&type=txt"

Next.js app router

bash

curl -sS "https://context7.com/api/v2/libs/search?libraryName=nextjs&query=app%20router" | jq '.results[0].id'
curl -sS "https://context7.com/api/v2/context?libraryId=/vercel/next.js&query=app%20router&type=txt"

FastAPI dependency injection

bash

curl -sS "https://context7.com/api/v2/libs/search?libraryName=fastapi&query=dependency%20injection" | jq '.results[0].id'
curl -sS "https://context7.com/api/v2/context?libraryId=/fastapi/fastapi&query=dependency%20injection&type=txt"

Optional CLI helpers

If you want a single command to search/fetch, use the bundled scripts:

bash

./scripts/context7.py search react hooks
./scripts/context7.py fetch /websites/react_dev_reference useState

bash

./scripts/context7.sh search react hooks
CONTEXT7_TYPE=txt ./scripts/context7.sh fetch /websites/react_dev_reference useState

References

•Full endpoint details: references/API.md
•Troubleshooting: references/TROUBLESHOOTING.md
•Optional helper scripts: scripts/context7.sh, scripts/context7.py