Intenttext Mcp
Parse, validate, query, and render IntentText (.it) intent documents inside the agent so specs stay machine-checkable and human-readable.
Overview
io.github.intenttext/intenttext-mcp is a Build-phase MCP server that parses, validates, queries, and renders IntentText (.it) documents for AI coding agents.
What is this MCP server?
- Full lifecycle for IntentText (.it): parse, validate, query, and render
- npm @intenttext/mcp-server v1.0.1 with stdio transport for local agents
- Keeps agent edits against a explicit intent grammar instead of loose markdown
- GitHub source at intenttext/intenttext-mcp for auditing tool behavior
- Fits pipelines where requirements travel as .it files next to the repo
- Package version 1.0.1
- npm identifier @intenttext/mcp-server
What problem does it solve?
Agent-edited requirements in markdown silently diverge from what you intended and cannot be machine-validated.
Who is it for?
Solo builders standardizing on IntentText (.it) who want local stdio MCP tooling in Claude Code or Cursor.
Skip if: Teams with no .it documents or those who only need generic Markdown linting without an intent grammar.
What do I get? / Deliverables
Your agent can treat .it files as structured intent—checking validity and answering queries—before generating or changing code.
- Validated and queryable .it intent documents via MCP tools
- Rendered views of intent content for agent context
- Repeatable parse/validate loop on every agent session
Recommended MCP Servers
Journey fit
Build is where you formalize intent artifacts that agents and codegen follow; IntentText sits alongside code and docs rather than replacing market research in Idea. docs is the right shelf because .it files are structured intent documents—parse, validate, query, and render map to spec maintenance and agent-readable requirements.
How it compares
Intent-document MCP toolkit, not a general documentation CMS or test runner.
Common Questions / FAQ
Who is intenttext-mcp for?
Developers and solo founders using IntentText (.it) who want agents to validate and query intent files instead of guessing from prose.
When should I use intenttext-mcp?
Use it during Build whenever you author, refactor, or implement against .it specs and need the agent to parse and validate before coding.
How do I add intenttext-mcp to my agent?
Install @intenttext/mcp-server from npm, register the stdio MCP server in Claude Code or Cursor config pointing at that package, and open your .it files in the workspace.