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 kanyun-inc/reskill --skill vibe-coding-guardiangit clone --depth 1 https://github.com/kanyun-inc/reskillWrote 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/kanyun-inc/reskill/vibe-coding-guardian)<a href="https://agentmods.dev/skills/kanyun-inc/reskill/vibe-coding-guardian"><img src="https://agentmods.dev/badge/skills/kanyun-inc/reskill/vibe-coding-guardian/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/kanyun-inc/reskill/vibe-coding-guardian"><img src="https://agentmods.dev/badge/skills/kanyun-inc/reskill/vibe-coding-guardian.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Prompt Injection · line 106 Large whitespace padding was detected (a block of blank lines or a long run of spaces). This can push injected instructions below or to the right of the visible area so a human reviewer never sees them while the agent still reads them. Manual review of the hidden content is recommended.Fix: Remove the large whitespace padding (blank-line blocks or long space runs) and review any content hidden below or to the right of it. Keep skill files compact and reviewable so no instructions can be
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 | $0.00073 | $0.01693 |
| Opus 5.5 | $0.00029 | $0.00677 |
| Sonnet 5.5 | $0.00015 | $0.00339 |
| Haiku 4.5 | $0.00007 | $0.00169 |
Grade A, and why
vibe-coding-guardian 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.
How it starts
The opening of the file, as written. The whole thing — 175 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Vibe Coding Guardian
Make the AI fast when it should be fast, careful when it should be careful. Small changes go straight through, risky changes get explained first, errors get handled automatically when possible, and the user only gets interrupted when truly necessary.
This is a behavioral modifier — it changes how the AI interacts with the user, not what it analyzes. The goal is to prevent the two most common disasters non-developers face: debug death spirals and code that gets worse with every fix.
Language rule: Always match the user's language. If they write in Chinese, respond in Chinese. If in English, respond in English. All explanations, verification prompts, and error translations should be in the user's language.
Core Rules
1. Explain Changes by Risk Level
Not every change needs user confirmation. Adapt behavior based on risk:
Low risk (fix typo, adjust styling, add comment):
- Make the change directly
- Briefly mention what was done
Medium risk (add feature, change logic):
- Describe what you're about to change and why
- Then make the change immediately — do NOT wait for confirmation
- Guide verification after (Rule 3)
High risk (delete files, rewrite architecture, change core logic across multiple files):
- Describe what will change and what might break
- Wait for user confirmation before proceeding
The key distinction: "explain" does not mean "wait for permission." Most of the time, say what you're doing and do it. Only truly dangerous operations need a pause.
2. One Change at a Time
Never bundle unrelated changes. If the user asks for 3 features, implement them sequentially:
- Feature A → verify it works
- Feature B → verify it works
- Feature C → verify it works
If Feature B breaks Feature A, it's immediately obvious which change caused it.
3. Guide Verification After Changes
After making changes, tell the user exactly how to check if it worked. Be specific:
- "Refresh the page — you should see a moon icon in the top-right corner"
- "Right-click the extension icon → Options — a new settings panel should appear"
- "Run
npm startand open http://localhost:3000 — the login form should show up"
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 · 175 lines · 73 tokens per session scan A bc619e3e8004
vibe-coding-guardian is a skill published in the GitHub repository kanyun-inc/reskill (60 stars, last pushed 1mo ago), licensed MIT. It adds 73 tokens to every session and 1,693 once invoked, about $0.0003 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 skills, from other repositories
vibe-prd
Write or revise an MVP product requirements document with scope and observable acceptance criteria.
vibe-research
Research an app idea or resolve product and technical uncertainties before committing to a design.
vibe-techdesign
Write an MVP technical design from agreed requirements, including architecture choices and relevant tradeoffs.
vibe-workflow
Select the appropriate planning, implementation, or debugging workflow for an app project.
vibe-agents
Create project agent instructions and tool configuration from agreed product and technical decisions.
vibe-build
Implement a new-project slice through the relevant checks and working user journey.