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 ohm41321/luciazero --skill debuggit clone --depth 1 https://github.com/ohm41321/luciazeroWrote 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/ohm41321/luciazero/debug)<a href="https://agentmods.dev/skills/ohm41321/luciazero/debug"><img src="https://agentmods.dev/badge/skills/ohm41321/luciazero/debug/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/ohm41321/luciazero/debug"><img src="https://agentmods.dev/badge/skills/ohm41321/luciazero/debug.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 | $0.00057 | $0.00556 |
| Opus 5.5 | $0.00023 | $0.00222 |
| Sonnet 5.5 | $0.00011 | $0.00111 |
| Haiku 4.5 | $0.00006 | $0.00056 |
Grade A, and why
debug scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 15d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 59 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Debug
For bugs that resist the first obvious look: debugging starts with a hypothesis, not an edit. Reality, not a plausible patch, decides.
1. Reproduce first
Find one command that shows the failure deterministically. It is ground truth. Do not theorize about causes of a failure you cannot trigger.
For intermittent failures, fix the seed, pin the time/timezone, run it in a loop, and identify the trigger before proceeding.
2. Minimize
Shrink to smaller input, fewer flags, one test instead of the suite. Stop when further reduction costs more than it clarifies.
3. Hypothesis ledger
Before inventing causes, search the symptom in:
- repo ledger
docs/lessons.md; - configured harness
luciazero-heuristics.mdunder${CLAUDE_CONFIG_DIR:-$HOME/.claude}or${CODEX_HOME:-$HOME/.codex}.
A match becomes H1, but must pass its proven-by observation.
H<N>: <suspected cause> — refutable by: <command / observation> → <refuted|confirmed|pending>
Run the observation, not the edit. Choose the cheapest discriminating log, query, debugger stop, grep, or bisect. Trust real state. Dead hypotheses stay in the ledger marked refuted. Keep the ledger visible in the conversation.
4. One variable per iteration
Change one thing, re-run the reproduction, and record the result. A failed fix gets reverted before the next attempt. Two consecutive failed fixes on the same hypothesis means the hypothesis is dead; widen to environment, versions, input, concurrency, or a wrong test.
5. Close out
- Commit a regression test. Run it both ways and quote both results: red before
the fix, green after. Use the done skill's
revert-probe.cjswhen applicable; once the fix is committed, pass the pre-fix commit as its base ref. - Remove all instrumentation: prints, sleeps, debug flags.
- Run the full verify tier.
- If the cause, footgun, or null result is not obvious from code, run
/retro; debugged failures go todocs/lessons.mdas symptom → cause → proof → fix.
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.
- 15d ago Changed · +1 lines f262c3c4886b
- 1mo ago First seen · 58 lines · 57 tokens per session scan A 2c99b57c0721
debug is a skill published in the GitHub repository ohm41321/luciazero (6 stars, last pushed 13d ago), licensed MIT. It adds 57 tokens to every session and 556 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
systematic-debugging
4-phase root cause debugging: understand bugs before fixing.
build-test
Run the project's build / typecheck / lint / test commands and emit the build.passing + tests.passing signals devloop convergence reads.
resolve-author-fix
Find the root cause, implement a focused fix, and prove behavior with targeted tests and recordings.
error-analysis
Evals-first error analysis for LLM apps: clusters real Langfuse or JSONL traces into a human-confirmed failure taxonomy with counts, then recommends binary pass/fail evals for recurring named modes. Use to learn what to measure before writing evals. Not for CI failures.
paired-probe
Refuse a verdict a probe did not earn. Runs a check where the fault IS present and where it is NOT, and blocks the answer when both arms print the same thing, because a check that cannot disagree with you has measured nothing. Also catches the zero-sample sweep that reads as "clean" and the swallowed error that reads…
nw-bugfix
Bug fix workflow: root cause analysis → user review → regression test + fix via TDD.