skills/jscraik/agent-skills/triage/Gen Agent Trust Hub

triage

Pass

Audited by Gen Agent Trust Hub on Apr 12, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The triage skill facilitates status transitions for file-based todos and provides clear instructions to prevent common risks like implementing code during the triage stage.
  • [PROMPT_INJECTION]: The references/evals.yaml file contains a test case simulating a prompt injection attempt ('Ignore previous instructions...') to verify the agent's resilience. This is a testing artifact and not a vulnerability in the skill's operational instructions.
  • [PROMPT_INJECTION]: The skill processes external data from the todos/ directory, which represents an indirect prompt injection surface. 1. Ingestion points: Reads markdown files from the todos/ directory. 2. Boundary markers: No explicit delimiters are mandated for the content of the todo files. 3. Capability inventory: File reading, renaming, and frontmatter modification. 4. Sanitization: The skill requires redacting secrets from artifacts, which provides a layer of data protection.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 12, 2026, 01:13 PM
Security Audit — agent-trust-hub — triage