# Rust Debug

> Rust debugging workflow for reproducing failures, panics, compiler diagnostics, trait/borrow errors, flaky tests, async hangs, deadlocks, and targeted fixes with regression tests. Use when this capability is needed.

- Skill: `tomevault-io/rust-debug` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add tomevault-io/rust-debug`
- Raw SKILL.md: https://api.skillmd.com/api/skills/tomevault-io/rust-debug/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Productivity
- Author: tomevault-io (https://skillmd.com/u/tomevault-io)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/tomevault-io/rust-debug

---


# Rust Debug

## Rule

Reproduce first, identify root cause, then make the smallest targeted fix. Prefer compiler
diagnostics, tests, backtraces, and tracing evidence over broad rewrites.

## Hard Stops

Stop before:

- Touching live systems, real secrets, production data, or destructive operations.
- Adding `unsafe`, changing public APIs, or weakening error contracts solely to silence an
  error.
- Suppressing panics, Clippy findings, or test flakes without understanding root cause.
- Fixing async hangs with sleeps or unbounded timeouts.

## Workflow

1. Capture failing command, panic, compiler diagnostic, backtrace, log, race/deadlock symptom,
   or hang.
2. Reproduce with the narrowest `cargo test`, `cargo check`, or executable command.
3. Isolate root cause using `RUST_BACKTRACE=1`, targeted tests, Clippy, `dbg!`/temporary
   tracing, Miri for unsafe/suspicious UB, sanitizers when configured, or `tokio-console` for
   async runtime issues.
4. Add a regression test when practical.
5. Apply a targeted fix and remove temporary diagnostics.
6. Run the original failing command, targeted tests, full tests, Clippy, and `just check`.

## Completion

Report reproduction, root cause, fix, regression tests, commands, and remaining risks.

---
> Source: [nyquistwilder/personal-pi](https://github.com/nyquistwilder/personal-pi) — distributed by [TomeVault](https://tomevault.io).
<!-- tomevault:4.0:skill_md:2026-06-16 -->