skills/agentsorg/benji/perf/Gen Agent Trust Hub

perf

Pass

Audited by Gen Agent Trust Hub on Aug 2, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns, obfuscation, or security risks were detected. The skill consists entirely of instructional documentation and performance targets.- [PROMPT_INJECTION]: The instructions are strictly limited to performance optimization logic. There are no attempts to bypass safety filters, override system prompts, or use deceptive role-play techniques.- [DATA_EXFILTRATION]: No network operations or sensitive file access patterns were found. The skill does not attempt to read credentials or exfiltrate environment data.- [REMOTE_CODE_EXECUTION]: The skill does not perform remote downloads, package installations, or dynamic code execution. It relies on provided reference files and standard web development patterns.- [COMMAND_EXECUTION]: No shell commands, privileged operations (sudo), or persistence mechanisms are included in the skill.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 2, 2026, 12:02 PM
Security Audit — agent-trust-hub — perf