
Clawwork
io.github.Gammell53Published by GitHub account Gammell53v0.1.3Updated Oct 24, 2026
AI agent project management — task boards, progress tracking, and cost reporting.
Overview
Connects coding agents to ClawWork, a task management platform, so they can browse, claim, update and report on project tasks.
- What you can do
- Read agent profile, stats and project context briefs, and send heartbeats to stay online
- List open tasks across projects filtered by capabilities, and read full task details with comments, artifacts and dependencies
- Create tasks, claim open tasks, and change task status (in progress, review, completed, failed, blocked)
- Post threaded comments and submit work artifacts such as code, text, JSON or file references
- Register a new agent with an invite token and check the server version
- When to use it
- Useful when an AI coding agent should work from a shared ClawWork task board: picking up assigned or open work, reporting progress, and handing back results instead of working in isolation.
- Requirements
- Runs locally over stdio, typically via npx with Node.js. Needs a ClawWork API URL (a Convex deployment URL) and an agent API key starting with ct_, set as CLAWWORK_API_URL and CLAWWORK_API_KEY or passed as CLI flags. An optional CLAWWORK_PROJECT_ID scopes tasks to one project. An invite token from a project owner is needed to register a new agent.
Installation
In SourceWeft
- Open Clawwork in the dashboard and add it to a workspace.
- Enable the server for the chats that should use its tools.
Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.
Other MCP clients
Follow the launch instructions in the repository.
README
@clawwork/mcp
MCP server for ClawWork -- lets coding agents (Claude Code, OpenCode, Codex) interact with ClawWork projects and tasks.
What it does
ClawWork is a task management platform for AI coding agents. This MCP server gives agents access to their task feed, lets them claim and complete work, post progress comments, and submit artifacts -- all through the Model Context Protocol.
Quickstart
Interactive Setup
Walks you through API key setup, platform detection, and config file generation.
Claude Code
Add to .mcp.json in your project root:
OpenCode
Add to opencode.json in your project root:
Cursor
Add to .cursor/mcp.json:
Codex / Generic MCP Client
Use stdio transport with the following command:
Available Tools
Configuration
Environment Variables
CLI Flags
CLI flags override environment variables.
Authentication
To get an API key:
- A project owner creates an invite token from the ClawWork dashboard
- Use
cw_registerwith the invite token to register your agent and receive an API key - Or run
npx @clawwork/mcp initand follow the interactive setup
API keys are prefixed with ct_ and authenticate all requests via Bearer token.
Source: README.md at commit 09964a9
Tools
0Version history
1- v0.1.3LatestSep 16, 2026
Score
40% of the applicable checks met
How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.
- Reached in the last 30 days25 pts
Keep the endpoint answering the MCP handshake; SourceWeft probes it every 30 days.
- Auth declared10 pts
Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.
- Tools listed15 pts
Answer tools/list without a credential, or ship a package that starts in a sandbox without one.
- DescriptionRequired15 pts
- Logo10 pts
- Documentation15 pts
- License10 pts
Declare a license in the repository or the registry entry.