M
MCP Directory CURATED DIRECTORY
911 MCP servers indexed
中文
← Back to directory
C

ChimeDeck MCP Server

Community
Open-source work management platform with AI agent integration.
How to install ↓ Compare GitHub source repository ↗
Category
Collaboration #53 of 55
Stars
★ 113 Popular
Transport
stdio (local process)
Runtime
Bun 1.3.5+ · Docker
Credentials
API key / credential required
License
AGPL-3.0
Last commit
Tools
4
28FMRS · D

ChimeDeck is a feature-rich open-source work management platform, particularly suited for teams aiming to leverage AI agents for workflow automation. It offers comprehensive task management and AI integration via MCP. However, self-hosting requires technical expertise and infrastructure, and the project may still be actively developed.

Strongest · Maintenance 7/20 Weakest · Reliability 5/20

Reliability
5/20
Security and permissions
5/20
Maintenance
7/20
Documentation
6/20
Setup experience
5/20
Why each score
Reliability 5/20
The repository lacks a server manifest, so MCP initialization and tool list cannot be confirmed. README claims MCP integration, but no MCP server implementation or tool definitions are found in the source. Migration and test scripts exist, but CI results are not cached, and test coverage is unclear. Uncertainty lowers the score.
Security and permissions 5/20
No explicit red-line issues found, but credential management (e.g., .env, AWS credentials) details are opaque, and least-privilege explanations are missing. Sentry privacy defaults are documented, but MCP server security boundaries and confirmation mechanisms for dangerous operations are not specified.
Maintenance 7/20
Repository is active with commits and 2 open issues, but no release or dependency update records observed. AGPL-3.0 license present. Governance and release path are unclear.
Documentation 6/20
README is detailed but lacks MCP-specific tool parameters, authentication, limits, and troubleshooting documentation. Setup steps exist, but MCP configuration examples are absent. Documentation does not fully match capabilities.
Setup experience 5/20
Installation steps (Bun, Docker, migrations) are clear, but how to start and connect the MCP server to a client is not explicit. No ready-made client config snippet provided; platform compatibility and complexity unverified.

Static review · not runListed 2026-08-07

Read the FMRS scoring method →

Fit and risk

What it can accessWrites / deletes local filesUses the networkConnects to a database

Best for

  • Teams needing AI-assisted project management
  • Developers looking to integrate AI agents into workflows
  • Those seeking an open-source, self-hosted work management solution

Not for

  • Those requiring fully managed SaaS (contact development team)
  • Environments without Bun or Docker
  • Users unfamiliar with MCP protocol

Required permissions

  • Read and write workspace data
  • Create and modify tasks, cards, comments
  • Access search functionality

Risks and side effects

  • Data security: self-hosting requires own data protection measures
  • External dependencies: requires Postgres, S3, etc.
  • AI-generated code may have defects; requires manual review

Setup

Before you start

Runtime：Bun 1.3.5+ · Docker

DATABASE_URL requiredsecret Full Postgres connection string (e.g., RDS endpoint); required.
S3_BUCKET required AWS S3 bucket name for attachment storage; required.
AWS_ACCESS_KEY_ID requiredsecret IAM access key ID for S3 access; secret.
AWS_SECRET_ACCESS_KEY requiredsecret IAM secret access key for S3 access; secret.
AWS_REGION required AWS region, e.g., us-east-1; required.
SENTRY_SERVER_DSN optionalsecret Server Sentry DSN from your Sentry project settings; secret.
SENTRY_AUTH_TOKEN optionalsecret Sentry API token with project:releases scope; needed only for source-map upload at build time; secret.
VITE_SENTRY_CLIENT_DSN optionalsecret Client Sentry DSN from your Sentry project settings; secret.
Other optional settings (15)
FLAG_USE_REDIS optional Set true to enable the Redis adapter; optional.
FLAG_VIRUS_SCAN_ENABLED optional Set false to disable attachment virus scanning; optional.
AUTOMATION_USE_PGCRON optional true enables pg_cron scheduling; false uses Bun Worker fallback; optional.
SEED_TRELLO optional Set true for first deployment to import Trello data, then set back to false.
FEATURE_FLAGS_JSON_PATH optional Path to a feature-flags JSON file; optional.
FEATURE_FLAGS_PROVIDER optional Remote flag provider (Flagsmith / FeatBit); optional.
SENTRY_SERVER_ENABLED optional Enable server-side Sentry capture; default false; optional.
SENTRY_ENV optional Deployment environment tag; must match client and server.
SENTRY_RELEASE optional Release identifier (git SHA or semver tag); must match build.
SENTRY_ORG optional Sentry organisation slug; needed only for source-map upload.
SENTRY_PROJECT optional Sentry project slug; needed only for source-map upload.
VITE_SENTRY_CLIENT_ENABLED optional Enable React client Sentry capture; default false.
VITE_SENTRY_ENV optional Client deployment environment tag; must match server.
VITE_SENTRY_RELEASE optional Client release identifier; must match server.
VITE_SENTRY_REPLAY_ENABLED optional Opt-in client Session Replay; bandwidth-intensive; default off.
  1. Clone repository: git clone https://github.com/Chimedeck/chimedeck
  2. Install dependencies: bun install
  3. Copy environment config: cp .env.example .env
  4. Start Postgres and LocalStack: docker compose up -d postgres localstack
  5. Run database migrations: bun run db:migrate
  6. Start dev server: bun run dev:local
  7. Configure MCP client.

Check that it works

Start the server (bun run dev:local) and run curl http://localhost:3000/health — a response of { "status": "ok" } confirms it is running; then check your MCP client lists the create_task, update_card, search_work, and add_comment tools.

Troubleshooting

  1. Ensure Bun version ≥1.3.5
  2. Check Docker services are running
  3. Verify .env configuration
  4. Inspect server logs: bun run dev

Things to try

Once connected, you can ask your AI assistant things like:

  • Create a new task in Chimedeck titled 'Organize customer feedback'
  • Update this card's description with the latest progress
  • Search all work items related to the pricing tools
  • Add a comment to this card saying the design mockups are approved

Tools 4

create_task writes
Create a new task in ChimeDeck.
update_card writes
Update details of an existing card.
search_work read-only
Search for work items and related content.
add_comment writes
Add a comment to a card.

Use cases

AI agents automatically create and manage tasks
Team members update card status via AI assistants
Quickly search work items in projects
Automated comments and feedback

Supported clients

Claude Desktop

Listed from the project's documentation, not tested by this site.

Overview

ChimeDeck is a full-stack workspace and card management app featuring boards, lists, comments, attachments, pricing tools, CLI commands, and automated tests. It enables MCP-compatible AI agents to directly create tasks, update cards, search work, and add comments, facilitating collaboration with human teams. Built with React, Bun/TypeScript, and an extensible server/client architecture for rapid product iteration.

Similar servers

Compare with other MCP servers

Source revision ec9c55dbc722 Data synced 2026-10-12 Read the FMRS scoring method