debugduck AGENTS.md

Repository instructions for debugging software problems, including failures, flaky tests, crashes, slow behavior, memory leaks, and incorrect output. They require reproducing the problem and finding where observed behavior first differs from what was expected.

In plain words
What is it for?
Investigating bugs, CI-only failures, hangs, regressions, logging requests, and execution flows by following a reproduce-first, evidence-based debugging process.
Why use it?
They reduce guesswork and prevent premature fixes that make it harder to identify the real cause.

Instructions file for CodexOpenCode

Install

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.

agentmods
npx agentmods add instructions/ujjwal502/debugduck/agents-md
Clone the repo
git clone --depth 1 https://github.com/ujjwal502/debugduck

Made for: Codex, OpenCode.

Per session 692 This file is loaded in full into every session.
When invoked 692 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00692 $0.00692
Opus 5 $0.00346 $0.00346
Sonnet 5 $0.00138 $0.00138
Haiku 4.5 $0.00069 $0.00069

Measured 2d ago against content hash 2b63c44ab387, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

debugduck AGENTS.md 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 2d 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.

AGENTS.md · 45 lines

How it starts

The opening of the file, as written. The whole thing — 45 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AGENTS.md

Debugging

When anything is broken, failing, flaky, slow, crashing, or behaving unexpectedly — a bug report, stack trace, failing or flaky test, regression, hang, memory leak, wrong output, a CI-only failure, or "works on my machine" — follow the procedure in skills/debug/SKILL.md. Read it before editing any code, then load only the reference files under skills/debug/references/ that match the bug in front of you.

If asked to add logs, instrument a flow, or trace execution from A to Z, follow skills/debug/references/trace-first.md.

The short version, which the full skill enforces:

  1. Reproduce before you diagnose. Diagnose before you fix. No repro means no feedback signal. If you genuinely can't reproduce it, say so and switch to evidence-mining — don't fix on spec.
  2. The bug is where the evidence points, not where it would be convenient.
  3. One variable at a time. Simultaneous changes destroy causality.
  4. To localize, trace the whole path in one pass. Instrument every checkpoint from entry to symptom at once (numbered CP01…CPnn), write down what each should print, run once, and find the first checkpoint where actual ≠ expected — everything upstream is then ruled out. Then stop and hand over the run command; don't guess a fix while waiting for the output. Protocol: skills/debug/references/trace-first.md. Fall back to binary search — code path, git history (git bisect run), the input, the config — when you can't re-run, the path is enormous, or logging perturbs the bug.
  5. A root cause explains 100% of the observed evidence. Unexplained details mean you found a bug, maybe not the bug.
  6. Prove the fix by toggling it: revert it, watch the bug return; reapply it, watch it go. Then add a regression test and verify it fails on the pre-fix code.
  7. Never suppress a symptom you don't understand — no empty catch blocks, blind optional chaining, added retries, bumped timeouts, or # type: ignore presented as a fix.
  8. Label every claim [observed], [inferred], or [assumed], and keep a hypothesis ledger (hypothesis → prediction → cheapest discriminating test → result) rather than guessing repeatedly.

Read the full file on GitHub · 45 lines

Changes

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.

  1. 2d ago First seen · 45 lines · 692 tokens per session scan A 2b63c44ab387

Subscribe to this mod's changes

debugduck AGENTS.md is an instructions file published in the GitHub repository ujjwal502/debugduck (4 stars, last pushed 24d ago), licensed MIT. It adds 692 tokens to every session, about $0.0035 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.