docs: add AGENTS.md and llms.txt agent-memory layer (#199)

[w7-mgfcode]

Summary

Stands up the AI-agent memory-file layer recommended by the emerging cross-tool standards (closes #199). Four committed files:

• AGENTS.md (new) — the universal cross-tool agent brief (open AGENTS.md standard; read by Claude Code, Cursor, Copilot, Gemini CLI, Aider, Zed, …). Plain Markdown, no frontmatter. Holds the shared project context: overview, tech stack, setup, build/test commands, validation gates, architecture & conventions, testing requirements, safety, git/PR conventions, deep-dive doc index.
• CLAUDE.md (slimmed 116 → 55 lines) — now the thin Claude-specific layer: it @imports AGENTS.md, keeps the @docs/_base/*.md deep-dive imports, a Safety pointer, Verification, Workflow, and Learnings. No content is duplicated between the two files; shared material lives only in AGENTS.md.
• llms.txt (new) — repo-root index following the llms.txt convention: H1 + blockquote summary + H2-sectioned links into the docs tree + an ## Optional section. Relative paths (the repo has no hosted docs site). All links verified to resolve.
• llms-full.txt (new) — the core docs (README, AGENTS.md, CLAUDE.md, the docs/_base/ suite) inlined for full-context LLM consumption.

Verification

• wc -l CLAUDE.md → 55 lines (well under the 150 budget).
• AGENTS.md commands cross-checked against Makefile, pyproject.toml, frontend/package.json.
• All llms.txt links resolve to real paths.

Out of scope / notes

• Two local skills generate and maintain these files — managing-agent-memory (AGENTS.md + CLAUDE.md + the docs/_base/ suite; supersedes and replaces w7_generating-claudemd) and managing-llms-txt (llms.txt + llms-full.txt). .claude/ is gitignored, so those skills are local-only and not part of this PR.
• No Cursor/Copilot symlinks — single-maintainer repo; AGENTS.md only.
• A pre-existing working-tree uv.lock change (self-version bump) was deliberately left out of this commit to keep the PR docs-only.

Acceptance

• AGENTS.md created with repo-verified commands
• CLAUDE.md @imports AGENTS.md, no duplicated content, ≤150 lines
• llms.txt + llms-full.txt follow the spec format

[sourcery-ai]

Sorry @w7-mgfcode, you have reached your weekly rate limit of 500000 diff characters.

Please try again later or upgrade to continue using Sourcery

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on base/target branches other than the default branch.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: 96b46322-cf46-4c25-b3f9-72a0970e4ce2

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch docs/agents-md-and-llms-txt

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}