missions

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). The skill deliberately exposes repository contents and diffs to external agent runtimes while using explicit "dangerous" bypass flags, and it automatically merges agent-produced changes into main and hides local agent state from git — creating clear vectors for data exfiltration, remote code execution, and supply-chain/backdoor insertion.

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

• Third-party content exposure detected (high risk: 0.70). The workflow ingests outsider-authored free text via the user-provided mission goal/plan (and --plan file) copied into LLM prompt inputs (mission-brief.md), which are then read and injected into the orchestrator prompt at runtime; no such file is treated as trusted-authored by the user.