deep-dive

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly instructs the agent to read runtime and project config files and inject trace-derived text and codebase_context into prompts and saved spec artifacts (e.g., settings.json, config.jsonc, .agents/), which can contain API keys or secrets and therefore forces the LLM to handle or output secret values verbatim.