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 beingmartinbmc/jambavan --skill root-cause-debuggergit clone --depth 1 https://github.com/beingmartinbmc/jambavanWrote 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/beingmartinbmc/jambavan/root-cause-debugger)<a href="https://agentmods.dev/skills/beingmartinbmc/jambavan/root-cause-debugger"><img src="https://agentmods.dev/badge/skills/beingmartinbmc/jambavan/root-cause-debugger/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/beingmartinbmc/jambavan/root-cause-debugger"><img src="https://agentmods.dev/badge/skills/beingmartinbmc/jambavan/root-cause-debugger.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00055 | $0.00721 |
| Opus 5.5 | $0.00022 | $0.00288 |
| Sonnet 5.5 | $0.00011 | $0.00144 |
| Haiku 4.5 | $0.00006 | $0.00072 |
Grade A, and why
root-cause-debugger 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 1mo 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.
How it starts
The opening of the file, as written. The whole thing — 65 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Root Cause Debugger
If the root_cause MCP tool is available, call it with the symptom (and attempts_so_far if this is not the first try) and use its output as an investigation checklist. Otherwise, follow the protocol below directly.
Evidence before fixes
Gather enough evidence to explain the likely failure mechanism before proposing a durable fix. If an urgent containment step is needed, label it as mitigation rather than root-cause resolution. After three unsuccessful fixes, pause and reassess assumptions, reproduction quality, and design boundaries; repeated failures do not by themselves prove the architecture is wrong.
Phase 1: Observe (before touching anything)
- Read the error completely. Stack traces, line numbers, error codes — do not skim.
- Reproduce consistently. Can you trigger it reliably? What are the exact steps?
- Check recent changes. What changed?
git diff, recent commits, new deps, config. - Trace data flow. Where does the bad value originate? Trace backward, not forward.
At multi-component boundaries, add the smallest safe diagnostic instrumentation needed:
- Log what enters each component.
- Log what exits each component.
- Run to gather evidence showing where it breaks.
- Remove temporary instrumentation when it is no longer needed, and do not log secrets.
Phase 2: Compare
- Find a working example of similar code in this codebase.
- List every difference between working and broken — however small.
- Do not assume "that can't matter."
Phase 3: Hypothesize and Test
- Form ONE hypothesis: "X is the root cause because Y."
- Make the SMALLEST possible change to test it.
- One variable at a time.
- Did it work? → Phase 4. Did not? → form a new hypothesis rather than stacking unverified fixes.
Phase 4: Fix
- Where practical, write or identify a failing test that reproduces the bug.
- Fix the root cause (not the symptom).
- Verify: test passes, no other tests broken.
- If 3+ fixes have failed, pause and discuss the evidence and next investigative step with the human.
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.
- 1mo ago First seen · 65 lines · 55 tokens per session scan A 0d605aeba862
root-cause-debugger is a skill published in the GitHub repository beingmartinbmc/jambavan (2 stars, last pushed 7d ago), licensed MIT. It adds 55 tokens to every session and 721 once invoked, about $0.0002 per session on Opus 5.5. 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
diagnose-with-memory
A bug-diagnosis workflow that uses records of earlier failures, decisions, and lessons. It checks whether past findings still apply to the current version and environment before investigating the code.
safe-refactor
A small-step refactoring workflow that keeps existing behavior unchanged. Refactoring means improving code structure without intentionally changing what the software does.
fix-bug
A workflow for reproducing a software bug, identifying its cause, and applying the smallest fix. It includes focused verification and optional failure-first tests.
ccc
This skill should be used when code search is needed (whether explicitly requested or as part of completing a task), when indexing the codebase after changes, or when the user asks about ccc, cocoindex-code, or the codebase index. Trigger phrases include 'search the codebase', 'find code related to', 'update the…
pre-action-interceptors
Use when a tool error begins with "INTERCEPTOR:", a sent message was silently hedged or rewritten, or the user asks about active guards or blocked actions. Not for ordinary tool errors.
observability
Periodic health sweep of the running system via the debug pipeline (cloud logs / k8s / daemon / DB slices). On a real anomaly, open a bug task (--kind workitem) carrying the evidence refs so it re-enters the bugfix loop. A healthy sweep opens NO task — never fabricate a bug.