gather
Pass
Audited by Gen Agent Trust Hub on Feb 23, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- [PROMPT_INJECTION] (LOW): The skill establishes an indirect prompt injection surface by instructing the agent to ingest untrusted data and perform state-changing actions. Ingestion points: Conversation history and session insights. Boundary markers: Absent; no delimiters or warnings to ignore embedded instructions. Capability inventory: File-write operations (project notes, memory) and network-enabled Git operations (syncing). Sanitization: Absent; no validation of mined content.
- [NO_CODE] (SAFE): No scripts, binaries, or executable code files are included in this skill.
Audit Metadata