performance-evidence-log
Pass
Audited by Gen Agent Trust Hub on Nov 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from multiple sources including 1:1 history, meeting transcripts, updates, and feedback. Ingestion points: Data is retrieved via tools such as find-one-on-ones, find-meetings, and find-feedback as defined in SKILL.md. Boundary markers: The skill includes explicit directives to treat retrieved text as untrusted and to ignore embedded instructions. Capability inventory: The skill can search report structures, gather performance data, and initiate feedback requests through the Lattice MCP. Sanitization: Instructions require human review for all employment-related outputs and explicit user confirmation before any write operation.
Audit Metadata