grill-me

Pass

Audited by Gen Agent Trust Hub on May 2, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it is instructed to investigate local project files and code to minimize redundant questioning.
  • Ingestion points: The skill reads project files and code directly into the agent context (SKILL.md).
  • Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present for the ingested data.
  • Capability inventory: The skill possesses filesystem write capabilities to create directories and logs.
  • Sanitization: No sanitization or validation of the content from the investigated files is performed.
  • [COMMAND_EXECUTION]: The skill performs filesystem operations by creating a .grill/ directory and writing markdown session logs to it (SKILL.md).
Audit Metadata
Risk Level
SAFE
Analyzed
May 2, 2026, 08:51 AM