supabase-extract-jwt

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly instructs the agent to extract JWTs from client-side code and to progressively write findings and evidence to context/evidence files (including decoded claims and examples showing raw token strings), which requires the LLM to handle and likely output secret token values and sensitive claims verbatim.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 1.00). This skill fetches and analyzes client-side code from arbitrary target web applications (see "Extract JWTs from https://myapp.example.com" and references to /static/js/* files), so it ingests untrusted public third-party content which the agent is expected to read and interpret.