perf-theory-gatherer

Pass

Audited by Gen Agent Trust Hub on Mar 9, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill requires the ability to execute system commands such as git and grep to gather performance data and history from the repository.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8).
  • Ingestion points: Processes untrusted data from git history, repository maps, and file content retrieved via grep or direct file access.
  • Boundary markers: No explicit delimiters or instructions to ignore embedded commands within the analyzed repository data are defined.
  • Capability inventory: Relies on git and grep tools which interact with the local filesystem and repository metadata.
  • Sanitization: There is no evidence of content sanitization or validation for the data retrieved from the repository before it is interpreted by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 9, 2026, 05:12 AM