skills.rest
eumemic
eumemicCOMMUNITY
Explore all skills by this author →
GitHub

bughunt

Audit codebase risk patterns and prove bugs with failing red tests.

Updated Feb 23, 2026
Search Tags:#testing#tdd#root cause#bug hunting#codebase audit#autonomous loop#github ci
One-click install
npx skills add https://github.com/eumemic/dev-skills --skill bughunt-eumemic
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: bughunt
Source: https://github.com/eumemic/dev-skills/tree/main/skills/bughunt
Command: npx skills add https://github.com/eumemic/dev-skills --skill bughunt-eumemic

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you continuously uncover real, reachable bugs by auditing code for risk patterns and then proving each candidate with a failing red test before fixing at the root cause.

Core Features & Use Cases

  • Parallel risk-pattern audits: Scans the codebase across invariant, concurrency, lifecycle, type-safety, error handling, boundary cases, derived state, time handling, query construction, idempotency, and coverage gaps to generate ranked bug hypotheses.
  • Provability-first TDD loop: Requires a specific, reachable hypothesis that can be proven by a failing test and flipped to green by a root-cause fix.
  • Quality-gated shipping workflow: Hands off to /ship once the bug is verified, runs /retro, and either merges automatically (with CI green) or waits for user merge.

Quick Start

Ask the assistant to run /bughunt --once to identify one provable, reachable bug hypothesis, fix it with a red-to-green test, and initiate the ship flow.

Frequently Asked Questions about bughunt

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

FAQPage Schema
How do I find real reachable bugs in my codebase using TDD?▼

Finding reachable bugs with TDD requires auditing codebase invariants and risk patterns to generate testable hypotheses, then proving each candidate with a failing red test before applying a root-cause fix to turn it green.

What codebase risk patterns should I audit to find hidden bugs?▼

Codebase risk pattern audits should scan for concurrency, lifecycle, type-safety, error handling, boundary cases, derived state, time handling, query construction, and idempotency issues to generate ranked bug hypotheses.

Can I automate bug hunting and fixing in my GitHub CI workflow?▼

You can automate bug hunting by running an autonomous loop that identifies provable hypotheses, fixes them with red-to-green tests, and hands off to a shipping workflow that merges automatically when CI is green.

What is the best way to debug and fix root causes instead of patching symptoms?▼

The best way to fix root causes is to require a specific, reachable hypothesis proven by a failing test, then apply a root-cause fix to flip it green, silently skipping any unprovable or unproducible hypotheses.

Why does my bug hunting process skip certain unprovable bug hypotheses?▼

Bug hunting processes skip unprovable hypotheses to maintain quality-gated shipping, ensuring only reachable bugs validated by failing red tests and root-cause fixes proceed to the ship flow.

Related Skills

Explore More Software Engineering Skills→
joekhosbayarjoekhosbayar

gds-test-review

Analyze test results and coverage to generate a structured test-review report.

Community
Intermediate
ajsai47ajsai47
11

land-and-deploy

Merge GitHub pull requests and verify production health after deployment.

Community
Advanced
MrlykMrlyk
4

test-driven-development

Enforce failing tests before implementation with per-repo test commands and trace.jsonl logging.

Community
Intermediate
SocketDevSocketDev
125

scanning-vulns

Scan source files for static vulnerabilities and output VULN-FINDINGS.json and Markdown.

Official
Advanced