gateguard

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

• Third-party content exposure detected (high risk: 0.75). The skill’s required workflow forces Claude to run Grep/Glob/Read to enumerate imports and file contents (e.g., “List ALL files that import/require this file (use Grep)” and “If this file reads/writes data files, show field names… (use redacted or synthetic values)”), which ingests repository text authored by others (outsider code/docs) into the LLM context via the local codebase read path.