plan-runner

Pass

Audited by Gen Agent Trust Hub on Jun 18, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection because its workflow involves interpreting and executing tasks from potentially untrusted external documentation.
  • Ingestion points: Plan files are read as the primary source of truth in the SKILL.md workflow.
  • Boundary markers: No specific delimiters or instructions are used to distinguish plan content from agent instructions.
  • Capability inventory: The agent is empowered to modify local files and execute validation scripts based on the plan's contents as described in SKILL.md.
  • Sanitization: No sanitization or validation of the plan file content is mentioned before the agent begins execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 18, 2026, 02:14 AM
Security Audit — agent-trust-hub — plan-runner