instrument-events
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes
event_candidatesYAML, which is untrusted data that influences the agent's behavior. - Ingestion points: The
event_candidatesYAML input (SKILL.md, Section 1). - Boundary markers: The instructions lack explicit delimiters or "ignore embedded instructions" warnings for the YAML data.
- Capability inventory: The agent is directed to read local files based on paths found in the YAML (
filefield) and output a JSON tracking plan. - Sanitization: There is no requirement to validate or sanitize the file paths or content within the
event_candidatesYAML before the agent performs file reads.
Audit Metadata