ducky

A debugging mode that only asks questions, like talking through a problem with a person who listens. It limits each response to two questions so you work out the cause or solution yourself.

In plain words
What is it for?
Use it for debugging practice, learning, and interview preparation.
Why use it?
It helps you reason through a bug or learn a topic without receiving the answer immediately.

Skill for Claude CodeCodex

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 skills/repowise-dev/modpack/ducky
Any agent
npx skills add repowise-dev/modpack --skill ducky
Clone the repo
git clone --depth 1 https://github.com/repowise-dev/modpack

Made for: Claude Code, Codex.

Per session 62 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 539 The whole file, excluding the scripts and references it only reads on demand.
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 $0.00062 $0.00539
Opus 5 $0.00031 $0.00269
Sonnet 5 $0.00012 $0.00108
Haiku 4.5 $0.00006 $0.00054

Measured 2d ago against content hash d6a254e0ded6, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

ducky 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.

skills/ducky/SKILL.md · 48 lines

What it actually says

ducky mode

Activation: /ducky, "activate ducky", "ducky mode". Deactivation: /default, "deactivate", "normal mode".

Rules

  1. Questions only. Every response is one or two questions. No answers, no hints disguised as statements, no "have you considered…" leading questions that contain the answer.
  2. Max 2 questions per turn. Usually 1. Two only when they're genuinely independent.
  3. Open before closed. Prefer "What happens when X runs?" over "Does X return null?" Closed questions only when narrowing in.
  4. Build on the user's last answer. Each question advances the chain.
  5. No preamble. No "Great, let's think about this…". Just the question.
  6. Silence is allowed. If the user is mid-thought, a single short prompt ("And then?") is fine.

Question ladder (rough order)

  1. What's the observed behavior?
  2. What did you expect?
  3. What's the smallest input that reproduces it?
  4. What does do with that input?
  5. What assumption are you making about ?
  6. How could you verify that assumption?

Override

If the user says "just tell me", "give me the answer", "stop asking", "I give up": break mode immediately, answer directly, do not re-engage ducky until reactivated.

Boundaries

  • Code execution / tool use: still allowed, but only to gather info you'll then ask the user about. Never to fix without permission.
  • Safety-critical bug (data loss, prod) → break mode, warn directly, ask if they want to continue ducky
  • Factual lookups the user explicitly requests ("what does this API return?") → answer the fact, then return to questions

Edge cases

  • User answers wrong → don't correct. Ask a question that exposes the gap.
  • User stuck for 3+ rounds with no progress → offer escape hatch: "Want a hint, or keep going?"
  • Task is execution, not debugging → ducky is wrong tool, suggest /default

Token note

Tiny responses. Net token use depends on conversation length.

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 · 48 lines · 62 tokens per session scan A d6a254e0ded6

Subscribe to this mod's changes

ducky is a skill published in the GitHub repository repowise-dev/modpack (16 stars, last pushed 4mo ago), licensed MIT. It adds 62 tokens to every session and 539 once invoked, about $0.0003 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 2026-08-30.

Related

Other skills, from other repositories

verify-pr

This skill should be used to run a sandboxed deep verification of a qwen-code PR — "/verify-pr ", "深度验证这个 PR", A/B load-bearing proof against the base build, mock-free harnesses with wire oracles, and targeted gates — producing tmp/pr -verify- /report.md plus a machine-readable verdict. Designed for the token-free CI…

QwenLM/qwen-code · 89 tokens

codegraph

Analyze indexed codebases via graph database (neug) and vector index (zvec). Covers call graphs, dependencies, dead code, hotspots, module coupling, architecture reports, semantic search, impact analysis, bug root cause from GitHub issues, class diagrams (UML), and PR review (risk scoring, conflict detection…

QwenLM/qwen-code · 0 tokens

stuck

Diagnose frozen, stuck, or slow Qwen Code sessions on this machine. Scans for problematic processes, high CPU/memory usage, hung subprocesses, and debug logs. Use /stuck or /stuck to focus on a specific process.

QwenLM/qwen-code · 57 tokens

e2e-testing

Guide for running end-to-end tests of the Qwen Code CLI, including headless mode, MCP server testing, and API traffic inspection. Use this skill whenever you need to verify CLI behavior with real model calls, reproduce user-reported bugs end-to-end, test MCP tool integrations, or inspect raw API request/response…

QwenLM/qwen-code · 94 tokens

qwen-code-claw

Use Qwen Code as a Code Agent for code understanding, project generation, features, bug fixes, refactoring, and various programming tasks.

QwenLM/qwen-code · 33 tokens

structured-debugging

Hypothesis-driven debugging methodology for hard bugs. Use this skill whenever you're investigating non-trivial bugs, unexpected behavior, flaky tests, or tracing issues through complex systems. Activate proactively when debugging requires more than a quick glance — especially when the first attempt at a fix didn't…

QwenLM/qwen-code · 85 tokens