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.
npx skills add HumanDealer/vibeproof --skill debug-root-causegit clone --depth 1 https://github.com/HumanDealer/vibeproofWrote 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.
[](https://agentmods.dev/skills/humandealer/vibeproof/debug-root-cause)<a href="https://agentmods.dev/skills/humandealer/vibeproof/debug-root-cause"><img src="https://agentmods.dev/badge/skills/humandealer/vibeproof/debug-root-cause.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00034 | $0.00468 |
| Opus 5 | $0.00017 | $0.00234 |
| Sonnet 5 | $0.00007 | $0.00094 |
| Haiku 4.5 | $0.00003 | $0.00047 |
Grade A, and why
debug-root-cause 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 7d 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.
What it actually says
Debug to Root Cause
The Iron Law: no fix without a confirmed root cause. Patching a symptom you don't understand just moves the bug. Work the four phases in order.
When to use
Any error, test failure, 500, stack trace, "it worked yesterday", or behavior that doesn't match expectations. Use this instead of immediately editing code.
The four phases
1. Investigate (gather facts, change nothing)
- Reproduce it reliably. If you can't reproduce it, you can't fix it.
- Read the actual error message and stack trace — top to bottom.
- Capture the exact inputs, environment, and steps that trigger it.
- Don't theorize yet. Collect evidence.
2. Analyze (find the why)
- Trace from the symptom back to the source. Follow the data.
- Ask "why" repeatedly until you reach a cause, not a coincidence.
- Distinguish where it surfaces from why it happens — often different.
3. Hypothesize (one testable theory)
- State a single, specific hypothesis about the root cause.
- Predict what you'd see if it's true. Test that prediction.
- Wrong? Back to phase 2 with what you learned. Don't guess-and-check edits.
4. Fix (root cause + regression test)
- Fix the actual cause, not the symptom.
- Write a test that fails before the fix and passes after. This proves the cause and prevents regression.
- Run the full quality gate (rules/code-quality.md).
After the fix
Record it: add a gotcha entry to rules/error-prevention.md (what happened,
why, how to prevent). If it caused real damage, write a lesson into memory
(rules/memory-system.md). That's how the bug stays dead.
Anti-patterns
- Changing things randomly to see what helps.
- Adding a try/except that hides the error instead of fixing it.
- "Fixed" without reproducing the original failure first.
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.
- 7d ago First seen · 51 lines · 34 tokens per session scan A b6426190ab98
debug-root-cause is a skill published in the GitHub repository HumanDealer/vibeproof (2 stars, last pushed 3mo ago), licensed MIT. It adds 34 tokens to every session and 468 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 2026-08-31.
Other skills, from other repositories
browser-trace
Capture a full DevTools-protocol trace of any browser automation — CDP firehose, screenshots, and DOM dumps — then bisect the stream into per-page searchable buckets. Use when the user wants to debug a failed run, audit network/console/DOM activity, attach a trace to an in-progress session, or feed structured per-page…
ontoly-software-graph
Use Ontoly's deterministic Software Graph and MCP capabilities for repository architecture, request tracing, dependency analysis, configuration lookup, and impact analysis before falling back to source search.
infrastructure-doctor
Skill for the doctor infrastructure module providing repository-level diagnostics and safe, reversible automated repair. Use when the local checkout is acting up, before opening a bug, when onboarding a fresh clone, or as a periodic health probe. Every mutation is backed up under .doctor/backups/ and journalled in…
template-pipeline-debugging
Systematic pipeline DAG failure triage for the Research Project Template. USE WHEN ./run.sh or executepipeline.py fails, a stage stalls (setup, tests, analysis, render, validate, LLM, copy), pytest/coverage gate fails mid-pipeline, PDF render or validate breaks, Project Analysis finishes too fast with no figures, or…
auto-service-technician
Use when a task needs senior auto-service-technician judgment — diagnosing a check-engine-light or drivability complaint from DTCs and live data, deciding whether a code-named part actually needs replacing or the fault lies upstream, scoping a flat-rate repair order and labor time, choosing warranty vs. customer-pay…
commercial-industrial-electronics-repairer
Use when a task needs the judgment of a Commercial and Industrial Electronics Repairer — diagnosing a VFD fault trip on a production line, tracing a PLC I/O fault back to the sensor, the wiring, or the ladder logic, selecting arc-flash PPE for live-panel diagnostic work, triaging multiple simultaneous down-lines by…