Project-wide guidance for AI coding agents and contributors.
- This file applies to the entire repository.
- If a deeper directory adds its own
AGENTS.md, the deeper file overrides this one for that subtree.
- Keep local-dev reliability high (
DNS + TLS proxy + local CA). - Prefer safe defaults for local security.
- Preserve maintainable structure over quick one-off fixes.
- Format and checks before commit:
cargo run -p xtask -- fmt-checkcargo test -qcargo check --workspace
- Do not bypass failing checks unless explicitly requested by the maintainer.
- Keep changes scoped to one concern per commit.
- After branch work is complete and before merge, run the
health-checkskill (skills/health-check/SKILL.md).
- Use project tooling only:
cargo run -p xtask -- fmtcargo run -p xtask -- fmt-check
pre-commitis check-only. It blocks commits when formatting is dirty.
- Follow
docs/commit-message.md. - All commit message content must be written in English.
- Scale commit message size to change size:
- small changes: short subject (+ optional 1-2 bullets)
- medium changes: subject + up to 3 bullets
- large changes: full structured format in
docs/commit-message.md
- Prefer small, explicit changes.
- Avoid broad refactors unless requested.
- Keep behavior changes documented in
README.mdwhen user-facing. - Add/adjust tests for changed logic when practical.
src/ca.rs(private keys, trust install)src/dns.rsandsrc/server.rs(resolver forwarding, concurrency)src/proxy.rsandsrc/tls.rs(routing, body limits, TLS behavior)
When touching these paths:
- explain the threat/risk being addressed,
- keep failure modes explicit in logs,
- avoid weakening existing safeguards.
- Keep responses concise and factual.
- If uncertain, state assumptions clearly before changing behavior.