write-file`/`kill`; JSON on stdout, `exec` streams remote output + propagates the remote exit code; secrets only via env/stdin, never argv — tests substitute the interpreter via `SANDBOX_E2B_PYTHON
distribution, category breakdown, or timeline as fits), and the reviewer note carried over. In Cowork this renders inline. In Claude Code I'll write an HTML file to [outputs folder
distribution, category breakdown, or timeline as fits), and the reviewer note carried over. In Cowork this renders inline. In Claude Code I'll write an HTML file to [outputs folder
distribution, category breakdown, or timeline as fits), and the reviewer note carried over. In Cowork this renders inline. In Claude Code I'll write an HTML file to [outputs folder
distribution, category breakdown, or timeline as fits), and the reviewer note carried over. In Cowork this renders inline. In Claude Code I'll write an HTML file to [outputs folder
distribution, category breakdown, or timeline as fits), and the reviewer note carried over. In Cowork this renders inline. In Claude Code I'll write an HTML file to [outputs folder
distribution, category breakdown, or timeline as fits), and the reviewer note carried over. In Cowork this renders inline. In Claude Code I'll write an HTML file to [outputs folder
distribution, category breakdown, or timeline as fits), and the reviewer note carried over. In Cowork this renders inline. In Claude Code I'll write an HTML file to [outputs folder
distribution, category breakdown, or timeline as fits), and the reviewer note carried over. In Cowork this renders inline. In Claude Code I'll write an HTML file to [outputs folder
distribution, category breakdown, or timeline as fits), and the reviewer note carried over. In Cowork this renders inline. In Claude Code I'll write an HTML file to [outputs folder
distribution, category breakdown, or timeline as fits), and the reviewer note carried over. In Cowork this renders inline. In Claude Code I'll write an HTML file to [outputs folder
distribution, category breakdown, or timeline as fits), and the reviewer note carried over. In Cowork this renders inline. In Claude Code I'll write an HTML file to [outputs folder
distribution, category breakdown, or timeline as fits), and the reviewer note carried over. In Cowork this renders inline. In Claude Code I'll write an HTML file to [outputs folder
without explicit user approval.
Changes to this module require careful review of the safety invariants documented in the code comments.
#### Core concepts
- **`HeapReader<'a, T>`** — A scoped borrow
code in this repo is named, typed, structured, formatted, tested, and reviewed. Pure Python coding standard; framework-agnostic.
- **`documents/ai/guides/serena.md`** — Always-on. When and how to use Serena MCP for symbol
halves:
- **Interactive skills** (`.claude/skills/`) — read and write files in the
repo (no target-code execution, except `/dnr-hunt` / `/dnr-respond`, which
run the bundled demo app locally — see `docs/security.md`), run in this
A file Claude Code reads at the start of every session. It holds the commands, conventions and warnings the agent needs for this project.
Where does it go?
At the repository root. Claude Code also reads CLAUDE.md files in subdirectories when it works there.
What should it contain?
Build and test commands, the project's layout, conventions that aren't obvious from the code, and mistakes to avoid. Short files tend to work better than long ones.
CLAUDE.md or AGENTS.md?
Claude Code reads CLAUDE.md; most other agents read AGENTS.md. Many projects keep one and point the other at it.