xc MCP
MCP server for xc — the Markdown-based task runner that defines tasks in README.md files. Enables AI agents to discover and run xc tasks defined in a project's Markdown files, list available tasks, and integrate xc-based project workflows into agent-driven development. Xc is a popular tool that makes tasks self-documenting by embedding them directly in project README files.
Score Breakdown
⚙ Agent Friendliness
🔒 Security
Executes shell commands from README tasks. Review task definitions. No credentials. Local execution only. Careful with destructive operations.
⚡ Reliability
Best When
A project uses xc for task running and an AI agent needs to understand and execute project tasks — xc's README-based approach makes tasks self-documenting and easy for agents to discover.
Avoid When
Your project uses a different task runner (Make, npm scripts, Taskfile) — use the appropriate MCP for your actual task runner.
Use Cases
- • Running project tasks defined in README.md from development automation agents
- • Listing available xc tasks for a project from project coordination agents
- • Executing build, test, and deployment tasks from CI/CD agents
- • Integrating xc workflows into AI-assisted development from coding agents
- • Discovering project capabilities from README task definitions from onboarding agents
Not For
- • Projects not using xc as their task runner (use Make, npm scripts, or Taskfile MCPs for those)
- • Complex orchestration beyond what xc tasks define
- • Teams unfamiliar with xc's Markdown task format
Interface
Authentication
No authentication required — local task execution. xc must be installed. Reads README.md in the current project directory.
Pricing
Free open source MCP for xc task runner. Both xc and this MCP are free.
Agent Metadata
Known Gotchas
- ⚠ xc must be installed separately — `brew install joerdav/xc/xc` on macOS
- ⚠ Task execution runs shell commands — ensure agent has appropriate project permissions
- ⚠ Tasks are discovered from README.md in the working directory — must be run in project root
- ⚠ Destructive tasks (deploy, reset, clean) should be confirmed before AI agent execution
- ⚠ Community MCP targeting niche xc user base — well-implemented but small ecosystem
Alternatives
Full Evaluation Report
Detailed scoring breakdown, competitive positioning, security analysis, and improvement recommendations for xc MCP.
Scores are editorial opinions as of 2026-03-06.