spec-to-code-compliance

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill explicitly requires quoting exact evidence from documentation and code (code_excerpt with file+line numbers and "Every claim must quote original text"), which will force the model to reproduce any embedded secrets or credentials present in the provided artifacts verbatim.

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

• Third-party content exposure detected (medium risk: 0.65). SKILL.md requires “Documentation Discovery” and “Universal Format Normalization” of user-provided specification documents (Phase 0–1), which at runtime can include outsider-authored free text from attachments/URLs/messages the operating user chooses to supply; the LLM then ingests that text into the spec_corpus/Spec-IR for later alignment.