debug

debug is a skill for Claude Code, Codex from christyjacob4/claude-tricks. It costs 9 tokens per session (1,491 once invoked), scanned A, original, MIT.

A process for investigating and fixing failing tests, build errors, or runtime problems in software.

In plain words
What is it for?
Use it to diagnose failures, inspect likely causes in recent changes and history, and apply a verified fix.
Why use it?
It replaces guesswork with reproduction, recent-change checks, and review of error details until the issue is fixed and tests pass.

Skill for Claude CodeCodex

Part of the claude-tricks plugin — 10 skills, 1 agent shipped together

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/christyjacob4/claude-tricks/debug
Any agent
npx skills add christyjacob4/claude-tricks --skill debug
Clone the repo
git clone --depth 1 https://github.com/christyjacob4/claude-tricks

Made for: Claude Code, Codex.

Or install claude-tricks, the plugin that ships this one along with the rest of its 10 skills, 1 agent.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for debug

README.md
[![agentmods](https://agentmods.dev/badge/skills/christyjacob4/claude-tricks/debug.svg)](https://agentmods.dev/skills/christyjacob4/claude-tricks/debug)
Your own site
<a href="https://agentmods.dev/skills/christyjacob4/claude-tricks/debug"><img src="https://agentmods.dev/badge/skills/christyjacob4/claude-tricks/debug.svg" alt="Measured on agentmods" height="20"></a>
Per session 9 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,491 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00009 $0.01491
Opus 5 $0.00005 $0.00745
Sonnet 5 $0.00002 $0.00298
Haiku 4.5 $0.00001 $0.00149

Measured 4d ago against content hash a2b414a35744, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

debug scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/debug/SKILL.md · 219 lines

How it starts

The opening of the file, as written. The whole thing — 219 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Debug and Fix

Systematically debug and fix failing tests, build errors, or runtime issues using aggressive parallelization at every phase.

DO NOT STOP UNTIL THE ISSUE IS FIXED AND TESTS PASS.

Arguments

  • $ARGUMENTS - Test name, error message, or description of the issue

Phase 1: Reproduce and Gather Context

Launch three parallel agents simultaneously to maximize information gathering speed.

Agent 1 - Reproduce the failure:

If test failure:

# Run specific test
./gradlew test --tests "*$ARGUMENTS*" --info

# Or run all tests to find failures
./gradlew test

If build error:

./gradlew build --stacktrace

If runtime error, get the full stack trace and identify the failing component.

Capture all error details: full error message, stack trace, test name and class, file and line number, input that caused the failure.

Agent 2 - Check recent changes:

git diff HEAD~5 --name-only
git diff HEAD~5 -- <relevant paths>

Identify what files changed recently that could be related to the failure. Summarize which changes are most likely to have introduced the bug.

Agent 3 - Check git history:

git log --oneline -15
git log --oneline -5 -- <files related to $ARGUMENTS>

Determine if this is a regression by finding when the relevant code last changed and who changed it.

Wait for all three agents to complete. Synthesize their findings to determine:

  • Is this a single test failure or multiple?
  • Is it flaky (intermittent)?
  • Is it a regression (worked before)?
  • What is the scope of the problem?

Phase 2: Root Cause Analysis

Using the error details from Phase 1, launch three parallel agents to analyze all relevant code simultaneously.

Agent 1 - Analyze the failing test (use Explore or read directly):

Read and understand the failing test file completely. Document:

  • What the test expects
  • What setup and mocks it uses
  • What assertions are failing and why
  • Whether the test itself is correct or buggy

Read the full file on GitHub · 219 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 4d ago First seen · 219 lines · 9 tokens per session scan A a2b414a35744

Subscribe to this mod's changes

debug is a skill published in the GitHub repository christyjacob4/claude-tricks (2 stars, last pushed 5mo ago), licensed MIT. It adds 9 tokens to every session and 1,491 once invoked, about $0.0000 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other skills, from other repositories

platform-skills

Use when troubleshooting, implementing, reviewing, or auditing platform infrastructure as a system — where Kubernetes, GitOps, CI/CD, and security concerns intersect. Provides structured diagnosis with blast radius, validation steps, and rollback plan for: Kubernetes, Flux CD, Argo CD, Terraform, GitHub Actions…

nitinjain999/platform-skills · 137 tokens

la-chancla

Use only when the user explicitly requests a strict review of one named task, change, plan, release, or invariant. Never self-selected, never auto-loaded, never invoked by another skill or by the mexican-mom router. NOT a session-wide mode and NOT a warning counter.

mcasillas17/mexican-mom · 61 tokens

mexican-mom

Use only when the user explicitly asks for mexican-mom, mom, or la jefa — routes to the right discipline skill or shows the index.

mcasillas17/mexican-mom · 35 tokens

a-ver-ensename

Use before claiming your own work is fixed, passing, complete, compatible, secure, built, or deployed — the claim requires artifact output produced since the latest change. NOT for unverified external library or API facts; use cadena-de-whatsapp. NOT for claiming a repo artifact is absent; use y-si-lo-encuentro-que.

mcasillas17/mexican-mom · 75 tokens

ahorita-es-ahorita

Use when one task must preempt all other work — the user explicitly says stop, a production outage, a leaked or committed secret, an active security incident, a red build blocking the team, a broken critical path. NOT for time estimates, TODOs, or deferrals; use ahorita.

mcasillas17/mexican-mom · 69 tokens

frijoles-en-el-tupper

Use when writing or reviewing names — functions, variables, types, files, schemas, API fields, config keys — and the comments attached to them, to check that each label predicts what is actually inside. NOT for whether failure paths are handled; use pero-ponte-sueter.

mcasillas17/mexican-mom · 65 tokens