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 soleimanmansouri/debug-bank --skill auto-instrumentgit clone --depth 1 https://github.com/soleimanmansouri/debug-bankWrote 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/soleimanmansouri/debug-bank/auto-instrument)<a href="https://agentmods.dev/skills/soleimanmansouri/debug-bank/auto-instrument"><img src="https://agentmods.dev/badge/skills/soleimanmansouri/debug-bank/auto-instrument/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/soleimanmansouri/debug-bank/auto-instrument"><img src="https://agentmods.dev/badge/skills/soleimanmansouri/debug-bank/auto-instrument.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 | $0.00019 | $0.00808 |
| Opus 5 | $0.00010 | $0.00404 |
| Sonnet 5 | $0.00004 | $0.00162 |
| Haiku 4.5 | $0.00002 | $0.00081 |
Grade A, and why
auto-instrument scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 99 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Auto-Instrument Skill
Use when the symptom classifier returns no match and the domain catalog is empty. Do NOT skip Step 0.
Step 0: Confirm No Match
Verify the classifier was already run. Expected output: "No pattern match" or all checklist scores 0/3. If you haven't run the classifier yet, do that first.
Step 1: Generate Hypotheses
H1: [cause] — Test: [what log value would confirm this]
H2: [cause] — Test: [what log value would confirm this]
H3: [cause] — Test: [what log value would confirm this]
Rank by proximity to the error site. Max 5 hypotheses.
Step 2: Instrument
Ensure the log directory exists: mkdir -p .debug-bank
Python
# region DEBUG-H1
import logging as _dbg; _dbg.basicConfig(filename='.debug-bank/debug.log', level=_dbg.DEBUG)
_dbg.debug(f"[DEBUG H1] var={var!r}")
# endregion DEBUG-H1
JavaScript / TypeScript
// #region DEBUG-H1
const _fs = require('fs'); _fs.appendFileSync('.debug-bank/debug.log', `[DEBUG H1] var=${JSON.stringify(var)}\n`);
// #endregion DEBUG-H1
Go
// #region DEBUG-H1
_f, _ := os.OpenFile(".debug-bank/debug.log", os.O_APPEND|os.O_CREATE|os.O_WRONLY, 0644)
fmt.Fprintf(_f, "[DEBUG H1] var=%v\n", var)
_f.Close()
// #endregion DEBUG-H1
Rules: log ONLY to .debug-bank/debug.log, never stdout. Never modify business logic.
Step 3: Reproduce
Run the failing operation. Logs capture evidence per hypothesis.
Step 4: Analyze
Read .debug-bank/debug.log. For each hypothesis:
H1: CONFIRMED / REJECTED / INCONCLUSIVE — [evidence]
Step 5: Fix
Minimal fix for the confirmed hypothesis only. Verify it resolves the symptom.
Step 6: Extract Pattern Candidate
candidate_id: P-candidate-YYYY-MM-DD
name: [from root cause]
category: [config / data / async / auth / type / state / network]
symptom_keywords: [from symptom]
checklist:
- [confirmed hypothesis — what to check first]
- [log expression that revealed it]
- [signal that ruled out other hypotheses]
debugger_strategy:
breakpoints: [file:function where bug was found]
watch_expressions: [log expression that showed the evidence]
fix_summary: [one sentence]
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.
- 8d ago First seen · 99 lines · 0 tokens per session scan A 3abf76468126
auto-instrument is a skill published in the GitHub repository soleimanmansouri/debug-bank (1 stars, last pushed 3mo ago), licensed MIT. It adds 19 tokens to every session and 808 once invoked, about $0.0001 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
code-context
How to answer codebase questions with the code-context MCP tools (search, sql, reindex): ranked hybrid keyword+semantic search, relevance-ranked SQL aggregation over the index, and index lifecycle. Use when a question spans many files ("how does X work", "where is Y handled"), when ranking or counting code by topic…
support-ticket-triage
Triage support tickets by pulling helpdesk context, searching issue trackers and knowledge bases in parallel, classifying the issue, and giving reps a recommended next action.
recovery
Handle blocked verification outcomes.
fix-issue
Fix a GitHub issue end-to-end. Use when given an issue number to fix, implementing bug fixes, or addressing reported problems.
lint-and-fix
Run linters and fix violations, formatting errors, or style mismatches using Trunk. Use when code quality checks fail, before submitting PRs, or to repair "broken" linting states.
less-is-more
Architecture-first, reduction-first workflow for code changes, debugging, refactoring, and cleanup. Use to trace the owning path, prefer no edit and subtraction before additions, replace obsolete behavior instead of layering around it, keep scope closed, refute your own diagnosis and fix before believing them, verify…