Real-time global intelligence dashboard. AI-powered news aggregation, geopolitical monitoring, and infrastructure tracking in a unified situational awareness interface
MCP server for pkg.go.dev package docs
This repo provides an MCP server that searches pkg.go.dev and returns Go package documentation to an agent. It runs over stdio, so you connect it through your MCP client and use it as a lookup tool while coding or researching.
Builders who use an MCP-capable agent and want Go package docs on demand.
You can ask your agent to find Go packages and fetch their docs without switching to a browser.
What it does
Search pkg.go.dev
Find Go packages from pkg.go.dev through an MCP tool.
Fetch package info
Return package documentation and related details for a selected package or subpackage.
Local cache
Cache search results locally so the server does not search every time.
Release workflow
Builds and publishes binaries for multiple platforms with GitHub Actions.
How to get it
- 1Or you can install it from source, compile it by yourself
go install github.com/yikakia/godoc-mcp-server/cmd/godoc-mcp-server@latest
README
godoc-mcp-server
search golang packages and their docs from pkg.go.dev, provide the infomation to llm as mcp server
Install
You can download the compiled binary from the releases page.
Or you can install it from source, compile it by yourself:
go install github.com/yikakia/godoc-mcp-server/cmd/godoc-mcp-server@latest
Usage
just use your client to request. it servers on stdio
Todo
- localCache
- use a local cache to avoid search every time(From reddit user)
- searchPackage
- imported by how many packages
- getPackageInfo
- fixme: subpackage need new adapter
- get examples
- release
- use github actions to release for multiple platforms
Develop Experience
The description is very important, when describing the param should tell the client how to use one tool's output and change the format to match another tool's input.
For example, the output of searchPackage contains subpackage's name, but do not contain
the package's name. So if want the llm to use getPackageInfo to get the subpackage's info,
I wrote the description for param packageName about how to use it to work with searchPackage:
package name for search. if use searchPackages before, and user want to get the subpackage info. you should plus them for example, when user query mcp, and it return packageName: github.com/mark3labs/mcp-go/mcp and subpackage client, then if user want to get the client package info, you should set the packageName to github.com/mark3labs/mcp-go/mcp/client rather than client
When using this detail description, the llm will tell you some subpackage's name after
search. You can just tell it which package you want to get info, or which package and the
subpackage's name, it will combine them and call getPackageInfo to get the info.
So maybe we should make the description configurable by user, to make the tool more useful and efficiency.
Library Usage
The exported Go API of this module should currently be considered unstable, and subject to breaking changes. In the future, we may offer stability; please file an issue if there is a use case where this would be valuable.
License
This project is licensed under the terms of the MIT open source license. Please refer to MIT for the full terms.
Files in the repo
- .github
- cmd
- pkg
- .gitignore
- go.mod
- go.sum
- LICENSE
- README.md
Discussion (0)
Ask about usage, or say what you built with itSign in to join the discussion.
No comments yet. Be the first to say what this is good for.
More connectors
High-performance code intelligence MCP server. Indexes codebases into a persistent knowledge graph — average repo in milliseconds. 158 languages, sub-ms queries, 99% fewer tokens. Single static binary, zero dependencies.

Universal provider proxy for OpenAI Codex & Claude Code — use any LLM (Claude, Gemini, Grok, DeepSeek, Ollama…) with Codex CLI, App, SDK, and Claude Code
Git-native persistent memory for AI coding agents. Implements Google OKF v0.2 with sub-300µs in-memory BM25 search, embedded MCP server, and progressive disclosure. Slashes token bloat by 80% with zero external databases or dependencies. Built in pure Go.
Local-first code intelligence graph for MCP and CLI. Builds a persistent map of your codebase so AI coding tools read only what matters, with benchmarked context reductions on reviews and large-repo workflows.
Stop your AI from making things up — it proposes, deterministic tools decide, every claim checked against ground truth with evidence. Grounded facts and context survive resets. Reverse engineering is the proving ground. MCP server + CLI.