project-setup

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). The skill intentionally persists detected project context to an external "Engram" (via mem_context / mem_save) and can migrate project docs, which creates a real risk of unauthorized data exfiltration of repository context and metadata without an explicit per-run consent in automatic mode.

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

• Third-party content exposure detected (medium risk: 0.30). SKILL.md “Step 1 — Project detection” reads real files from the current project at runtime (package.json/README/docs, etc.) that could contain outsider-authored free text, and those contents are then used to generate CLAUDE.md and ai-context/.