AI and LLM security review
SkillSecurityOnce added, your AI can assess AI applications and language models for security weaknesses, covering risks like prompt injection, data leakage, and jailbreaks. It also helps review agent permissions, model supply chain concerns, and governance or evaluation questions. This makes it easier to spot and fix AI-specific security issues before they cause problems.
Available today. Use it from your connected AI after setup.
No other account needed.
After adding it, tell your AI which AI application or model setup you want reviewed and which risks matter most to you. It will then carry out the relevant assessments and summarize what it finds.
Then ask your AI: use the AI and LLM security review skill
What your AI can do with it
- Run security assessments and red teaming tests on AI and language model applications
- Test for prompt injection and jailbreak weaknesses
- Review retrieval setups for data leakage risks
- Check agent and tool permission settings
- Assess model supply chain risks
- Design evaluations and support AI governance reviews
What this skill tells your AI
The instructions your AI receives, as published by 26zl/cybersec-toolkit in .claude/skills/ai-llm-security-review/SKILL.md and read by ahel’s review.
Use this skill for AI applications, agents, RAG systems, model gateways, prompt chains, evals, and LLM governance.
Review workflow
- Inventory the AI system: model/provider, prompts, tools, RAG sources, memory, logs, user roles, secrets, data classes, and downstream actions.
- Threat model trust boundaries:
- user input to prompt
- retrieved content to model
- model output to tools
- tool output to user
- logs/traces to operators
- Test high-risk paths:
- direct and indirect prompt injection
- data exfiltration from RAG or memory
- insecure tool invocation
- overbroad agent permissions
- jailbreaks that change policy or role
- model/provider key leakage
- training/eval data contamination
- Recommend controls:
- least-privilege tool scopes
- allowlisted tool schemas and argument validation
- retrieval filtering and source attribution
- secret redaction before prompts/logs
- output validation before side effects
- human approval for destructive or external actions
- continuous evals and regression prompts
Deliverables
Return findings as:
| Risk | Attack path | Impact | Evidence | Control | Test to keep fixed |
|---|
When the task involves current AI regulation or sector obligations, verify against current official sources before making definitive claims.
Signals
- GitHub stars
- 54
- Forks
- 10
- Last commit
- Sep 2026
Advanced
- Catalog kind
- skill
- Gateway key
ai-llm-security-review- Source
- github.com/26zl/cybersec-toolkit