feature-design-doc

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill requires preserving and quoting user-provided code, logs, and source text verbatim and harvesting from session/code/diffs—actions that force the model to reproduce any embedded secrets (API keys, tokens, passwords) in its output if present.

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

• Third-party content exposure detected (medium risk: 0.30). The skill “feature-design-doc” ingests and reads free-text from the user/session conversation (and potentially a user-provided existing doc path) during Phase 1/3/4 to generate and fact-check FDD content, which is outsider-authored input processed directly by the LLM.