Is the Semantic Trap Detector skill free?

Yes. The skill itself is free from Jumbo-WJB/semantic-trap-detector. SkillProof publishes the install command and an independent test verdict at no cost.

What is the Semantic Trap Detector SkillProof Score?

8.0/10 — installs cleanly 5/5, triggers reliably 5/5, output vs. baseline 8/10, docs & honesty 2/5.

How do I install Semantic Trap Detector?

Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.

Can I use Semantic Trap Detector with Cursor, Copilot, Gemini CLI, Codex or other AI tools?

The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.

Semantic Trap Detector

Meta-skill that scans a SKILL.md's wording for words with over-wide semantic boundaries.

by Jumbo-WJB · Jumbo-WJB/semantic-trap-detector

Tested · Works

Install — copy the command Download source ↗

Test report

Verdict: Tested · Works
Score: 8.0/10
Tested: Jul 14, 2026
Environment: Claude Code 2.x (agent harness)
Upstream re-checked: Aug 10, 2026 · 8ce061f

Ran its own workflow on a test sentence and it correctly flagged 分析/风险/建议 with trap IDs and concrete replacements — genuinely more actionable than an unstructured critique — but the docs claim '17+ Chinese word pairs' while the actual lexicon.md file ships only 13, and the headline '89.3% vs 62.1%, 27-point' accuracy stat cites an unpublished 56-case test set with no data or link anywhere in the repo.

Scored on four weighted criteria — install, triggering, output vs. baseline, docs. How scoring works

Installs cleanly 5/5
Triggers reliably 5/5
Output vs. baseline 8/10
Docs & honesty 2/5

What Semantic Trap Detector does

A meta-skill for auditing other Claude Skills: it runs a fixed 4-step workflow (load lexicon, extract keyword categories, two-layer dictionary+context matching, structured JSON report) against a lexicon of Chinese/English word pairs (e.g. 风险 vs 漏洞, Analyze vs Summarize) that it claims cause LLM scope creep. Triggers on requests to test/optimize a skill's prompt wording or fix hallucination-prone instructions.

How to install Semantic Trap Detector

git clone https://github.com/Jumbo-WJB/semantic-trap-detector
cd semantic-trap-detector
mkdir -p ~/.claude/skills
cp -r . ~/.claude/skills/semantic-trap-detector

Skills live in ~/.claude/skills/ (global) or .claude/skills/ (per-project). Restart Claude Code after installing.

Commands — how to trigger Semantic Trap Detector

/semantic-trap-detector Meta-skill that scans a SKILL.md's wording for words with over-wide semantic boundaries.

It also activates on plain-language prompts like these:

Scan this SKILL.md wording for words that could cause scope creep
Does this prompt use vague terms an LLM might over-interpret?
Audit our skill instructions for overly broad verbs like 'analyze'

Frequently asked questions

Is the Semantic Trap Detector skill free?: Yes. The skill itself is free from Jumbo-WJB/semantic-trap-detector. SkillProof publishes the install command and an independent test verdict at no cost.
Does Semantic Trap Detector work with Claude Code?: We tested it with Claude Code 2.x (agent harness) on Jul 14, 2026. Verdict: Tested · Works. Ran its own workflow on a test sentence and it correctly flagged 分析/风险/建议 with trap IDs and concrete replacements — genuinely more actionable than an unstructured critique — but the docs claim '17+ Chinese word pairs' while the actual lexicon.md file ships only 13, and the headline '89.3% vs 62.1%, 27-point' accuracy stat cites an unpublished 56-case test set with no data or link anywhere in the repo.
What is the Semantic Trap Detector SkillProof Score?: 8.0/10 — installs cleanly 5/5, triggers reliably 5/5, output vs. baseline 8/10, docs & honesty 2/5.
How do I install Semantic Trap Detector?: Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.
Can I use Semantic Trap Detector with Cursor, Copilot, Gemini CLI, Codex or other AI tools?: The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.