WEEEK plugin with 12 curated tools
weeek-mcp
A Claude Code and Codex plugin for WEEEK: 12 curated, not auto-generated tools, with the token in the OS keychain
Medium risk
We rate an entry medium when the tool runs code, makes network calls or reads project files. Check what exactly it does before installing.
Why this level
- Creates, changes and completes tasks and comments in the WEEEK workspace
- The token is stored centrally in the OS keychain and available to every client on the machine
Install
In your terminal, with SkillFoxx CLI
npx skillfoxx add plugins/dsudomoin-weeek-mcpDetects the agents on your machine, checks the risk and pins the version.
Other ways to install
Run one by one in the Claude Code chat
/plugin marketplace add dsudomoin/weeek-mcp
/plugin install weeek@dsudomoinInstall as a plugin with claude plugin marketplace add dsudomoin/weeek-mcp and claude plugin install weeek@dsudomoin, then run weeek-mcp init in a terminal: the wizard asks for the token without showing it, checks it against WEEEK and stores it in the OS keychain.
Other ways from the author
claude plugin marketplace add dsudomoin/weeek-mcp
claude plugin install weeek@dsudomoinAfter installing, run weeek-mcp init to save the token to the keychain.
This is third-party code. Review the repository files before installing.
What it does
The server lets Claude Code and Codex work with WEEEK tasks in ordinary words: what is on the board, assign this to Anna and comment that the build is fixed, instead of switching to a browser. 12 tools cover 23 API operations: four reads (workspace context, task search across many filters, a task card with its description and comment thread, a standalone comment list for paging) and eight writes (creating and editing a task, moving it between boards and columns, completing or reopening, assigning members and watchers, comments, downloading attachments). The author deliberately curated and merged tools instead of auto-generating one per endpoint, so the list fits in every agent request. Answers come back in human-readable form: names instead of ids, markdown instead of HTML. The token lives in the OS keychain rather than a text file, and one install serves every client on the machine.
Who it is for. For people who want to work with WEEEK tasks in plain speech from Claude Code or Codex without the API's technical details getting in the way.
Good fit when
- You want a compact, curated set of WEEEK tools rather than a full auto-generated list
- You work offline, behind a corporate proxy, or want a pinned server version
- You want one installed token to work across every MCP client on the machine
Not a fit when
- You need WEEEK's CRM part (deals, funnels, contacts), this server is focused on tasks only
- You need to upload attachments to a task, removed as of 0.2.0, attachments can only be downloaded
Example request
What is on the Sprint 12 board, assign the Ship the release task to Anna, and comment that the build is fixedLimitations
Requires Node 22.18 or newer. Under Codex the token can only be set via the weeek-mcp init wizard, since the Codex plugin cannot ask for settings. As of a recent update, weeek_upload_attachment is removed along with WEEEK_FILE_ROOT; the server no longer uploads files or writes to a named path, attachments are added in WEEEK itself. On Linux without a session D-Bus, such as a container or SSH session, the token is stored in an owner-only-readable file instead of the OS keychain.
How to disable. Run claude plugin uninstall weeek@dsudomoin or codex plugin remove weeek@dsudomoin, and for a global install npm uninstall -g @dsudomoin/weeek-mcp.
Security check
- Creates, changes and completes tasks and comments in the WEEEK workspace
- The token is stored centrally in the OS keychain and available to every client on the machine
README in short
The English README explains the motivation for a curated tool set instead of auto-generating one per API endpoint, so the list fits every request. It shows installs via the Claude Code and Codex plugins, a global npm package for offline work, the weeek-mcp init wizard for secure keychain token storage, full tables of four read and eight write tools, and a separate section on upgrading from 0.1.0 and 0.2.0, including the removal of attachment uploads.
FAQ
Why only 12 tools instead of one per API method?
The author deliberately merged and curated them instead of auto-generating, so the tool list fits within every agent request's context.
Where is the token stored?
In the OS keychain after a one-time weeek-mcp init run, not in a plain-text config; on Linux without D-Bus the token goes to an owner-only-readable file.
Related
A self-hosted knowledge base with block-level references and a built-in MCP server for connecting AI agents to your notes
A CLI for every Google Workspace API with JSON output and agent skills: Drive, Gmail, Calendar, Sheets and more
Local search over Markdown notes, docs and meeting transcripts: keywords, semantic search and reranking, with an MCP server
A task manager for AI-driven development: breaks a PRD into dependent tasks and guides the agent through them via MCP or CLI