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.
git clone --depth 1 https://github.com/azrtydxb/procoderWrote 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/commands/azrtydxb/procoder/debug)<a href="https://agentmods.dev/commands/azrtydxb/procoder/debug"><img src="https://agentmods.dev/badge/commands/azrtydxb/procoder/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/commands/azrtydxb/procoder/debug"><img src="https://agentmods.dev/badge/commands/azrtydxb/procoder/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 | $0.00031 | $0.00616 |
| Opus 5.5 | $0.00012 | $0.00246 |
| Sonnet 5.5 | $0.00006 | $0.00123 |
| Haiku 4.5 | $0.00003 | $0.00062 |
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 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.
What it actually says
The user invoked /procoder:debug with arguments:
The command below is the procoder binary on PATH.
The iron law: no fixes without root-cause investigation first. Use this especially when a quick fix seems obvious, when you are under time pressure, or when the previous fix didn't work — systematic is faster than thrashing. The arguments (if any) describe the bug; start at phase 1 with them.
Phase 1 — root cause:
- Read the complete error: stack trace, line numbers, codes. They often contain the exact answer.
- Reproduce it consistently before touching anything. Not reproducible → gather more data, don't guess.
- Check what changed:
git log/diffrecent commits, new dependencies, config, environment. - Use the index to see the terrain:
procoder index callers <fn>andrefs <symbol>for every path into the failing code — the bug's home may be a caller, not the crash site. - For multi-component failures: log what enters and exits each boundary, run once to find WHERE it breaks, then investigate that component. Trace the bad value backward to its origin, not forward from the symptom.
Phase 2 — pattern: find a working example of the same thing in this codebase, read it completely (not skimmed), and list every difference — "that can't matter" is how the difference that matters gets skipped.
Phase 3 — hypothesis: state exactly one ("I think X because Y"), test it with the smallest possible change. Wrong → new hypothesis, never a second change stacked on the first. Don't know → say "I don't understand X" and investigate that; never say "this might work".
Phase 4 — fix: write the failing test FIRST (it pins the bug and proves the fix), then the single fix — no while-I'm-here refactoring. Verify: test passes, suite passes, gate clean.
The three-strikes rule: three failed fixes for the same bug is not bad luck, it is evidence the architecture is wrong — each fix revealing new coupling, or spawning new symptoms. STOP. Say so to the user, name the pattern that keeps failing, and discuss whether to refactor it before any fix number four.
Red flags that mean you are thrashing (return to phase 1): "quick fix for now, investigate later" · "just try changing X" · several changes then run the tests · "it's probably X" without evidence · fixes proposed before any investigation. And read the user's signals: "stop guessing" or "is that not happening?" means your evidence is missing, not their patience.
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 · 56 lines · 31 tokens per session scan A 478185e74aa5
debug is a command published in the GitHub repository azrtydxb/procoder (211 stars, last pushed 4d ago), licensed Apache-2.0. It adds 31 tokens to every session and 616 once invoked, about $0.0001 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 2026-08-30.
Other commands, from other repositories
logic-fix-all
Autonomous audit-and-fix — after consent, scan the target, fix every logic issue found (all severities), verify each fix, and report anything unresolved.
logic-explain
Trace execution step by step — use when code behavior is surprising or confusing.
logic-health
Sweep a whole codebase or directory for logic correctness — use before a release or to identify risk hotspots.
logic-locate
Locate the root cause of a confirmed failure — use when you have a stack trace, failing test, or wrong output.
logic-review
Review code for logic bugs — use when you suspect something is wrong but have no failing test yet.
verify
Verify code, architecture, or bug fixes using Chain-of-Verification.