@okfshare/mcp
Enables AI coding agents to search and retrieve cited line-level context from versioned knowledge bundles, as well as list workspaces and publish or pull immutable Markdown bundles.
README
okfshare CLI and MCP server
Client-side, open-source half of okfshare: shared, versioned knowledge for AI coding agents.
Publish a validated bundle of Markdown once. Every authorized agent works from the same immutable revision, with line-level citations you can audit. The hosted service keeps revisions, citations, and workspace auth; this repo holds the tools your agents actually run.
Packages
| Package | Command | Purpose |
|---|---|---|
okfshare |
npx okfshare@latest |
CLI: setup, publish, search, context, diff, rollback |
@okfshare/mcp |
npx @okfshare/mcp@latest |
MCP server for Claude Code, Cursor, opencode, Codex |
@okfshare/agent-installer |
okfshare-agent |
Installs okfshare skills into agent tooling |
Quick start
# 1. One-time onboarding (device pairing in your browser)
npx okfshare@latest setup
# 2. Validate and publish a folder of Markdown
npx okfshare@latest validate ./knowledge
npx okfshare@latest publish ./knowledge --visibility public --yes
# 3. Any teammate's agent retrieves cited context
npx okfshare@latest context SHARE_ID "how do we handle db migrations"
Headless/CI environments: supply a pre-provisioned credential via the
OKFSHARE_TOKEN environment variable (check existence only; never log it),
then verify with whoami and doctor.
MCP setup
claude mcp add okfshare -- npx @okfshare/mcp@latest
Or any MCP client:
{
"mcpServers": {
"okfshare": {
"command": "npx",
"args": ["@okfshare/mcp@latest"]
}
}
}
Tools: whoami, doctor, list_shares, search_share,
search_workspace, get_context, pull_bundle, publish_bundle.
Registry manifest: server.json.
Why citations matter
Retrieval returns the file, the exact lines, and the immutable revision they came from. When an agent claims "our convention says X", you can audit the claim in seconds - and a citation means the same thing tomorrow, because revision content never changes in place.
The Open Knowledge Format
Bundles are Markdown with structured frontmatter: a type, title,
description, tags, sources, and verification events. The format is
deliberately small and stays readable in any editor. Worst case, you are left
with plain files - no lock-in.
- Spec: https://github.com/GoogleCloudPlatform/knowledge-catalog/blob/main/okf/SPEC.md
- Field guide (published with okfshare itself): https://okfshare.app/s/ecf08367a22273482c0887ca8700ee8e
- Forkable conventions starter: https://okfshare.app/s/e8ddeda5aab619495c970f031dce7058
Development
pnpm install
pnpm check # typecheck + test + build across all packages
- Node 20+ and pnpm 10.
- The packages talk to the hosted API at
https://okfshare.app; no local server is required to develop them. - See CONTRIBUTING.md for the acceptance flow and SECURITY.md for reporting vulnerabilities.
Links
- Site and docs: https://okfshare.app
- Agent-readable docs: https://okfshare.app/llms.txt and https://okfshare.app/llms-full.txt
License
Apache-2.0. See NOTICE - the code license does not grant rights to the okfshare name or branding.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.