Getting started
What grāmatr is, the five-stage Loop in 60 seconds, and how to apply for access.
grāmatr documentation
Real-time intelligent context engineering. Five stages, one Loop, every request.
What you’ll find here
Section titled “What you’ll find here”MCP integration
Setup guides for Claude Code, Cursor, ChatGPT, Gemini, and any MCP-native tool.
API reference
Request shapes, response shapes, error codes, and stability guarantees for every endpoint.
Operations
Authentication, rate limits, and the governance scopes that gate cross-tier intelligence sharing.
Troubleshooting
Diagnostics for the common failure modes: token issues, CF Access, JWKS, intake errors.
Changelog
What shipped, what changed, and what behavior to expect across versions.
This site is a skeleton
Section titled “This site is a skeleton”We’re filling these sections in over the coming days. If something you need isn’t here yet, the source is at github.com/gramatr/gramatr-docs — file an issue or jump into the gramatr platform support channel.
For the marketing site (what grāmatr is, why we built it, the proof): gramatr.com.