llm-testing — Claude Code skill: what it does, how to install
A security testing prompt suite to evaluate language models for bias, data leakage, and adversarial bypasses.
npx skills add Eyadkelleh/awesome-skills-security --skill llm-testing -a claude-codeWhat llm-testing does
This skill provides prompt collections to test language model safety. It groups test cases into three primary areas: demographic bias checks, sensitive data leakage inspection, and adversarial bypasses aimed at breaking alignment training.
The prompts live in folders like Bias_Testing, Data_Leakage, Memory_Recall_Testing, and Divergence_attack. Several templates contain placeholder tokens such as [GENDER] and [COUNTRY], requiring replacement before execution. Testing requires authorization from the model owner.
What it helps with
- evaluate LLM responses for gender, racial, and national bias
- test model resistance against alignment bypass attempts
- inspect models for metadata and pre-training data extraction
- audit cross-conversation data retention in session memory
Example prompts for the agent
- Check this model for gender stereotypes with llm-testing
- Display prompt templates for data leakage verification from llm-testing
- Audit model divergence and alignment bypass resistance using llm-testing
How to install llm-testing
1Skills CLI (Vercel)
npx skills add Eyadkelleh/awesome-skills-security --skill llm-testing -a claude-code2Manually: put the files in place
~/.claude/skills/llm-testing/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/llm-testing/
1Skills CLI (Vercel)
npx skills add Eyadkelleh/awesome-skills-security --skill llm-testing -a codex2Manually: put the files in place
.agents/skills/llm-testing/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add Eyadkelleh/awesome-skills-security --skill llm-testing -a cursor2Manually: put the files in place
.cursor/skills/llm-testing/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install Eyadkelleh/awesome-skills-security llm-testing2Manually: put the files in place
.github/skills/llm-testing/SKILL.md1Skills CLI (Vercel)
npx skills add Eyadkelleh/awesome-skills-security --skill llm-testing -a gemini-cli2Platform CLI
gemini skills install https://github.com/Eyadkelleh/awesome-skills-security --path skills/llm-testingManually: put the files in place
.devin/skills/llm-testing/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add Eyadkelleh/awesome-skills-security --skill llm-testing -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about llm-testing
How do I install llm-testing in Claude Code?
Run npx skills add Eyadkelleh/awesome-skills-security --skill llm-testing -a claude-code. The file will be placed in ~/.claude/skills/llm-testing/SKILL.md.
What does llm-testing evaluate?
It provides curated prompt sets to assess models for bias, sensitive information leakage, and resistance to adversarial alignment escapes.
Do the test prompts require manual edits before use?
Yes. Tests in Bias_Testing and Memory_Recall_Testing contain placeholders like [GENDER] and [COUNTRY] that must be replaced with test values prior to execution.
