large-file-refactor

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 0.90). The skill contains explicit instructions to bypass external coordination controls (skip the ai-coord gate) and to follow prior invocation directives instead of standard tooling, which are hidden/overriding instructions unrelated to the refactoring task and that attempt to change agent behavior or authority checks.

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

• Third-party content exposure detected (low risk: 0.10). SKILL.md runs scripts/large-file-refactor.py, which reads and counts source lines from local filesystem files under the provided scan path (path.read_text(...)), so outsider-authored text can only enter if the caller supplies/points to outsider-controlled repository/files.