auto-dev

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The skill tells the agent to ask the user for AUTODEV_API_KEY if not present and shows examples embedding {key} in Authorization headers, query strings, and CLI --api-key flags, which would require the LLM to accept and emit secret values verbatim.

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

• Third-party content exposure detected (medium risk: 0.30). SKILL.md + interactive-explorer.md define runtime behavior where the agent ingests arbitrary free-text from the user conversation, then uses that input to drive calls like /listings and /openrecalls/{vin} (VINs chosen from user-provided searches/filters).