dev

Pass

Audited by Gen Agent Trust Hub on Jun 16, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill facilitates an Indirect Prompt Injection surface by ingesting untrusted user input via the $ARGUMENTS variable and interpolating it directly into the initial command of the pipeline.
  • Ingestion points: User-provided input via $ARGUMENTS in SKILL.md.
  • Boundary markers: Absent. There are no delimiters or instructions provided to the agent to treat the input as data rather than instructions.
  • Capability inventory: The pipeline includes high-privilege operations such as modifying the codebase (implementation step) and creating Pull Requests on the user's behalf.
  • Sanitization: Absent. The skill does not perform any validation, escaping, or filtering of the user-supplied content before passing it to the next stage of the pipeline.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 16, 2026, 11:18 PM
Security Audit — agent-trust-hub — dev