repomix-unmixer

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly parses and extracts arbitrary user-provided repomix files (XML/Markdown/JSON) — e.g., "/path/to/skills.xml" or "repo-output.xml" — ingesting their file contents for review/validation, which are untrusted/user-generated sources the agent will read and interpret.