Scan code for LLM invocations and generate compliance reports with remediation guidance.
npx skills add https://github.com/inernoro/prd_agent --skill llm-visibility
Please help me install this Agent Skill. Skill: llm-visibility Source: https://github.com/inernoro/prd_agent/tree/main/.claude/skills/llm-visibility Command: npx skills add https://github.com/inernoro/prd_agent --skill llm-visibility
LLM interactions are often opaque; this Skill visualizes and audits LLM invocation points in code to ensure compliance with a no-blank-wait policy and produce remediation guidance.
Scan your codebase for every LLM invocation and generate a visual audit report with remediation suggestions.
High-intent search queries and answers about installing and using this skill.
LLM invocation scanning checks code paths to locate calls and verify adherence to a no-blank-wait policy. It produces a compliance report identifying potential wait-time issues across streaming and non-streaming endpoints.
A no-blank-wait policy ensures LLM interactions maintain observable progress and utilize streaming outputs to prevent indefinite waiting. It also enforces safe data handling to mitigate data leakage risks during generation.
Detect prompt injection risks by scanning code to locate LLM invocations and evaluate them against safe data handling requirements. This identifies potential security vulnerabilities and provides remediation guidance to secure your endpoints.
Yes, you can scan both streaming and non-streaming endpoints for LLM wait-time issues. The scanning process works across backend and frontend code to identify potential wait-time violations and ensure observable progress.
Remediation guidance for LLM data leakage risks is generated by scanning code to locate LLM invocations and flagging non-compliant data handling. This produces a structured report with concrete steps to address the identified vulnerabilities.
msanssouciAutomate code reviews with backend and frontend convention checks.
mahg-esDesign and enforce CI/CD quality gates for code, tests, security, build, E2E, and performance.
ngquaanghuyAnalyze AI agent skill files for command injection, data exfiltration, and prompt injection.
uphiagoIdentify exploitable LLM/AI vulnerabilities crossing trust boundaries with validation gates.
