anvil-diagnose

Diagnose hard bugs and performance regressions through a six-phase feedback-loop workflow.

1|Updated Aug 27, 2026
One-click install
npx skills add https://github.com/rogerznts/anvil --skill anvil-diagnose-rogerznts
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: anvil-diagnose
Source: https://github.com/rogerznts/anvil/tree/main/anvil/.claude/skills/anvil-diagnose
Command: npx skills add https://github.com/rogerznts/anvil --skill anvil-diagnose-rogerznts

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve? Hard bugs and performance regressions resist casual code reading; without a reproducible signal, debugging devolves into guessing. This Skill enforces a disciplined diagnosis loop that builds a tight, deterministic pass/fail signal before any hypothesis is tested. ## Core Features & Use Cases - Feedback loop construction: Builds a red-capable, deterministic, fast, agent-runnable command (failing test, curl script, headless browser, replayed trace, bisection harness) that goes red on the exact bug. - Structured six-phase workflow: Reproduce and minimise, generate 3-5 ranked falsifiable hypotheses, instrument with tagged debug logs, fix with a regression test at the correct seam, then clean up all instrumentation. - Non-deterministic bug handling: Raises reproduction rates via looped triggers, parallelism, and stress until flaky bugs become debuggable. - Use Case: A user reports an intermittent export failure in production. The Skill guides building a replay harness from a captured request, minimising the repro, ranking hypotheses, and landing a fix with a regression test. ## Quick Start Ask the assistant to diagnose the bug where the export button throws an error, and have it build a failing reproduction loop before proposing any fix.

Frequently Asked Questions about anvil-diagnose

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I debug a bug that only happens sometimes?▼

For non-deterministic bugs, raise the reproduction rate instead of seeking a clean repro: loop the trigger 100 times, parallelise, add stress, and narrow timing windows. A 50%-flake bug is debuggable; keep increasing the rate until your feedback loop reliably goes red.

How to build a feedback loop for debugging a failing API endpoint?▼

Write a curl or HTTP script against a running dev server that asserts the exact symptom the user reported, not just absence of errors. Make it deterministic, fast (seconds), and runnable unattended so it can go red on the bug and green after the fix.

What should I do when I cannot reproduce a bug at all?▼

Stop and say so explicitly rather than hypothesising without a loop. Ask the user for environment access, a redacted captured artifact such as a HAR file or log dump, or permission to add temporary production instrumentation.

Why should the regression test be written before the fix?▼

Writing the test first at a correct seam proves it exercises the real bug pattern by watching it fail, then confirms the fix by watching it pass. If no correct seam exists, that absence is itself an architectural finding worth flagging.

How do I debug a performance regression differently from a functional bug?▼

For performance regressions, avoid log-based probing; establish a baseline measurement with a timing harness, profiler, or query plan, then bisect against it. Measure first, fix second, and verify against the original un-minimised scenario.