perf-profile

Pass

Audited by Gen Agent Trust Hub on Aug 24, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted data from the analyzed codebase.
  • Ingestion points: The skill uses Read, Glob, and Grep to extract data from repository files (SKILL.md).
  • Boundary markers: There are no instructions to the agent to disregard malicious content within those files.
  • Capability inventory: Access to the Bash tool is provided, which could be misused if injected instructions are followed (SKILL.md).
  • Sanitization: The skill lacks mechanisms to sanitize or validate the content it reads.
  • Self-referential claims: The text includes a self-assigned safety verdict ("Verdict: COMPLETE") meant to bypass external security evaluation (SKILL.md).
  • [COMMAND_EXECUTION]: The skill uses Bash for analysis tasks (SKILL.md). This introduces a risk of command injection when processing data sourced from untrusted files.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 24, 2026, 04:18 PM
Security Audit — agent-trust-hub — perf-profile