debug

A structured method for finding the root cause of bugs and errors. It gathers evidence, tests possible explanations, traces the failure backward through the code, and confirms the cause.

In plain words
What is it for?
Investigating failing tests, behavior changes after refactoring, production incidents, and errors by examining messages, inputs, expected results, timing, and execution paths.
Why use it?
It replaces guesswork with a repeatable investigation when a failure is unexpected or its source is unclear.

Skill for Claude CodeCodex

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/brain-bootstrap/claude-code-brain-bootstrap/debug
Any agent
npx skills add brain-bootstrap/claude-code-brain-bootstrap --skill debug
Clone the repo
git clone --depth 1 https://github.com/brain-bootstrap/claude-code-brain-bootstrap

Made for: Claude Code, Codex.

Per session 43 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 526 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 $0.00043 $0.00526
Opus 5 $0.00022 $0.00263
Sonnet 5 $0.00009 $0.00105
Haiku 4.5 $0.00004 $0.00053

Measured 2d ago against content hash f93eccae9309, 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 2d 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.

.claude/skills/debug/SKILL.md · 61 lines

How it starts

The opening of the file, as written. The whole thing — 61 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Debug Skill

Systematic root cause analysis.

When to use

  • An error is happening and the cause is not obvious
  • A test is failing unexpectedly
  • Behavior changed after a refactor
  • Production incident requiring root cause identification

5-Step Investigation Method

Step 1 — OBSERVE: Gather exact evidence

Do NOT guess. Collect all available evidence first:

  • Exact error message (full stack trace, not summary)
  • Exact input that triggers the failure
  • Exact expected vs. actual output
  • When it started failing (which commit? which change?)
  • Environment where it fails (dev? CI? prod? all?)

Step 2 — HYPOTHESIZE: Form concrete hypotheses

Based on the evidence, generate 2-3 specific hypotheses:

  • Each hypothesis must be falsifiable (testable)
  • Rank by probability based on evidence
  • Do NOT assume the most complex explanation — prefer simple causes

Step 3 — TRACE BACKWARDS: Follow the execution path

From the failure point, trace backwards:

grep -rn --color=never '<error_symbol_or_message>' . | head -20
  • Find the function that threw the error
  • Find its caller
  • Find the caller's caller
  • Keep tracing until you reach user input or a boundary

Step 4 — BINARY SEARCH: Narrow the scope

If the trace is long, use bisection:

  • Does it fail with the same input after reverting X? (git bisect or manual)
  • Does it fail with minimal input? (reduce the test case)
  • Does it fail in isolation? (extract the failing unit)

Step 5 — PROVE: Verify the root cause

Before fixing, prove you found the right root cause:

  • Write a test that reproduces the bug exactly
  • Verify the test fails (red) with the current code
  • Apply the fix
  • Verify the test passes (green)
  • Run the full test suite — no regressions

Rules

  • NEVER fix before proving the root cause
  • NEVER submit a fix without a reproducing test
  • If two hypotheses remain after Step 4, test the simpler one first
  • Document the root cause in claude/tasks/lessons.md if it's a non-obvious pattern

Read the full file on GitHub · 61 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. 2d ago First seen · 61 lines · 43 tokens per session scan A f93eccae9309

Subscribe to this mod's changes

debug is a skill published in the GitHub repository brain-bootstrap/claude-code-brain-bootstrap (11 stars, last pushed 4mo ago), licensed MIT. It adds 43 tokens to every session and 526 once invoked, about $0.0002 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 2026-08-30.

Related

Other skills, from other repositories

implement

TRIGGER when: user asks to implement, fix, build, or work on something — whether from a docs/wip plan OR a standalone task (bug fix, GitHub issue, one-off change). Examples: "work on task 1", "fix this bug", "implement feature X from the issue". Provides structured execution with profile detection, dependency…

serpro69/claude-toolbox · 79 tokens

review-spec

Use after implementing tasks or mid-feature to verify code matches design docs and ensure they are in sync. Detects spec deviations, missing implementations, doc inconsistencies, and outdated docs in design and implementation documentation.

serpro69/claude-toolbox · 44 tokens

chain-of-verification

Apply Chain-of-Verification (CoVe) prompting to improve response accuracy through self-verification. Use when complex questions require fact-checking, technical accuracy, or multi-step reasoning.

serpro69/claude-toolbox · 41 tokens

review-design

Review design, implementation, and task documents produced by design. Evaluates document quality, internal consistency, and technical soundness. Use after design completes and before starting implement.

serpro69/claude-toolbox · 37 tokens

review-code

Code review of current git changes with an expert senior-engineer lens. Detects SOLID violations, security risks, and proposes actionable improvements. Use when performing code reviews.

serpro69/claude-toolbox · 37 tokens

dependency-handling

TRIGGER when: adding or upgrading any dependency — library, SDK, framework, API, IaC API version (K8s/Terraform/Helm), CRD, or container image. Use BEFORE writing the call. Forces context7/capy lookup instead of guessing.

serpro69/claude-toolbox · 60 tokens