Context7 — up-to-date docs
A model knows a library as it was when the model was trained. Draw an arrow from an agent to the Context7 card and it looks up the current documentation and code examples instead — for the version your project uses.
Built on Context7, the open-source MCP server and docs index by Upstash:github.com/upstash/context7
2
tools an agent gets with one arrow: find the library, read its docs
24
hours an answer stays in the local cache — a repeated lookup costs no request
0
accounts needed — a free API key only raises the rate limit
One question, two tool calls, current docs
The calls and their results are the ones the real app passed to the model in a test run, against the live Context7 service.
❯ How do I clean up a React effect?
Ready
198 of 200 requests left
Recent lookups
Shortened for the page: the search returned 5 matching libraries and the docs call several examples. The agent’s answer is illustrative.
The two tools of Context7’s own server
The same tools and descriptions as Context7’s official MCP server, prefixed like every NeuroSquad plugin tool. NeuroSquad calls the Context7 service directly — nothing to install or start.
context7_resolve_library_idFinds a library by name and lists the matches: its Context7 id, how many code examples there are, how trusted the source is and which versions are indexed.
context7_query_docsReturns the documentation and code examples of one library for one focused question — for a specific version when the id names one.
Help before the agent even asks
Turn on Auto-docs on the card and NeuroSquad looks at each of your prompts before the agent starts.
A pointer, instantly
Off by defaultWhen the prompt names one of the project’s dependencies — from package.json, requirements.txt, pyproject.toml, Cargo.toml or go.mod — the agent gets a one-line note that current docs for it are available, with the version your project declares. No request is sent, so the turn starts just as fast.
The docs, when you say “use context7”
Add “use context7” to a prompt that names a dependency or a library id and the documentation itself is added to the turn. NeuroSquad waits about a second and a half; if the service is slower, the agent gets the pointer and the answer is cached for the next time.
Free to start, careful with requests
Context7’s documentation index is a hosted service with a rate limit. The card shows it and spends it carefully.
No account needed
It works without a key at a lower rate limit. The card shows the requests left and when the limit resets.
An optional free key
A free key from the Context7 dashboard raises the limit. It is stored encrypted on your computer and sent only in the request header — never in a file or a command line.
Rate-limited, clearly
When the limit is used up the card says so, agents get a plain message, and no further requests are sent until it resets.
A local cache
Every answer is kept on your computer for 24 hours. Two agents asking the same thing, or one agent asking again, cost one request.
One card, one arrow
Everything happens on the canvas you already have open.
Add the card
Add-card menu → “Plugin…” → Context7 — up-to-date docs → “Add to canvas”.
Draw an arrow
From an agent to the card. It gets the two tools on its next tool listing — no restart for most CLIs.
Work as usual
The agent checks the docs when a task touches a library API. You can also ask it to, or add “use context7” to the prompt.
Look at the card
See which library and question each agent looked up, whether it came from the cache, and how many requests are left. Try a lookup yourself.
Every CLI NeuroSquad runs
The tools come through NeuroSquad’s own MCP server, so every agent with an arrow can use them.
The two tools — every agent NeuroSquad connects to its MCP server
Auto-docs — agents whose turns NeuroSquad can add context to
What leaves your computer
Only what a lookup needs: a library name and the question about it.
Names and questions only
The library name and the question go to Context7’s servers, run by Upstash. Your code and files are never sent.
Agents are told
The tool descriptions tell agents never to put secrets, personal data or proprietary code into a question.
Dependencies read locally
Auto-docs reads your project’s dependency list on your computer and sends at most the one library name it matched.
A local option
Need lookups that never leave the computer? A self-hosted documentation MCP server can be added from the catalog’s MCP tab instead.
Open source inside
Powered by Context7
The tools, their descriptions and the requests are those of Context7’s official MCP server (@upstash/context7-mcp, MIT). The documentation comes from Context7’s hosted index, run by Upstash.
Questions about Context7
Agents that read the current docs
Download NeuroSquad, add the Context7 card, and draw an arrow.