Bootstrap Agent System

You are setting up this repository's AI coding agent operating system, modeled on a proven pattern. Port the system and doctrine, then rewrite all content for THIS repo based on evidence you gather here. Never paste product-specific details from another codebase unless they verifiably apply to this repo.

Goal

Give AI agents (Cursor, Claude Code, etc.) a single, living contract for how to work in this repo:

1. Root entrypoint docs (AGENTS.md + CLAUDE.md)
2. Root CONTEXT.md — the project's domain language (glossary + decisions)
3. Deeper agent docs under docs/agents/
4. Durable plans under docs/plans/
5. Release notes under docs/release.md
6. Project skills under the repo's skills directory (e.g. .agents/skills/)
7. Explicit rules that agents must use project scripts, keep docs living, and record lessons

Reference templates

This skill ships with templates in references/. Read each one when you reach the step that uses it: