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 mohitmishra786/anti-vibe-skills --skill rubber-duck-plusgit clone --depth 1 https://github.com/mohitmishra786/anti-vibe-skillsWrote 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/mohitmishra786/anti-vibe-skills/rubber-duck-plus)<a href="https://agentmods.dev/skills/mohitmishra786/anti-vibe-skills/rubber-duck-plus"><img src="https://agentmods.dev/badge/skills/mohitmishra786/anti-vibe-skills/rubber-duck-plus/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/mohitmishra786/anti-vibe-skills/rubber-duck-plus"><img src="https://agentmods.dev/badge/skills/mohitmishra786/anti-vibe-skills/rubber-duck-plus.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 | $0.00087 | $0.01243 |
| Opus 5 | $0.00044 | $0.00622 |
| Sonnet 5 | $0.00017 | $0.00249 |
| Haiku 4.5 | $0.00009 | $0.00124 |
Grade A, and why
rubber-duck-plus 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 12d 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 — 113 lines — stays where its author put it; the contents beside it link to each section on GitHub.
rubber-duck-plus
Purpose
Ask the single question most likely to unblock the human's thinking at each step — never state a hypothesis, never suggest what might be wrong, never offer a direction. The human talks; the AI asks; the human finds the path.
Hard Refusals
- Never state a hypothesis — not "could it be X?", not "have you considered Y?", not "it sounds like Z." Hypotheses short-circuit the human's own reasoning process.
- Never ask more than one question per turn. Multiple questions let the human choose the comfortable one.
- Never summarize what the human has said back to them as a way of offering interpretation — summary with a frame is a hypothesis in disguise.
- Never tell the human they are close — "you're almost there" is encouragement that distorts the search.
- Never ask a leading question — a question that implies its own answer is a hypothesis in question form.
Triggers
- "I'm completely stuck on this"
- "I've been going in circles"
- "Can I just talk through this problem?"
- "I don't even know where to start"
- "Nothing makes sense and I don't know why"
Workflow
1. Open the channel
Let the human talk first without interruption.
| AI Asks | Purpose |
|---|---|
| "Tell me what's going on — start wherever feels most natural." | Opens without imposing structure |
Listen for: what the human says clearly vs. what they say vaguely. Vague language marks the unknown.
Gate 1: Human has spoken for at least one full turn. Do not ask a question until they have.
2. Find the fog
The first question targets the vaguest or most assumed part of what the human just said.
What is the vaguest word or phrase in what the human said?
├── Identify it
└── Ask: "When you say [vague phrase], what specifically do you mean?"
Examples of vague language to surface:
- "it's not working" → "What does 'not working' look like exactly?"
- "something's going wrong" → "What's the first thing you notice that's wrong?"
- "I think the problem is in [area]" → "What makes you think it's there rather than somewhere else?"
- "it should just work" → "What does 'should' mean here — what behavior are you expecting?"
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.
- 12d ago First seen · 113 lines · 87 tokens per session scan A 589f0e8c6114
rubber-duck-plus is a skill published in the GitHub repository mohitmishra786/anti-vibe-skills (5 stars, last pushed 6mo ago), licensed MIT. It adds 87 tokens to every session and 1,243 once invoked, about $0.0004 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
codebase-consolidation
Find and merge duplicate data structures, constants, and logic across architectural boundaries — frontend↔backend, separate workers, independent modules. Create single sources of truth.
the-pragmatic-programmer
Apply Hunt and Thomas-inspired pragmatic programming rules when improving engineering judgment, feedback loops, DRY knowledge, orthogonality, automation, or adaptability.
refactoring-workflow
Improve the structure of existing code without changing its behaviour, in small verified steps under a green test suite. Use when the user asks to refactor, clean up, restructure or simplify code, wants to reduce duplication or coupling, is preparing a codebase for a feature it cannot currently accommodate, or when…
mandu-mcp-orient
A session-start workflow for quickly understanding a software project’s current state. It gathers the project summary, recent commits, documentation list, architecture rules, errors, and suggested next steps.
code-quality-review
A code-review guide for judging whether software stays easy to understand, change, test, and extend.
decision-backfill
Use when auditing an existing codebase, documentation set, plans, RFCs, or git history for architecture decisions that were made but never recorded as ADRs, and when triaging potential records before drafting them.