creating-effective-skills

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • Prompt Injection (SAFE): No instructions to bypass safety filters or ignore prior instructions were found. The skill uses standard instructional language to define its own behavior.
  • Data Exposure & Exfiltration (SAFE): No hardcoded credentials or sensitive file paths detected. The skill correctly advises on using specific, non-destructive tool permissions and warns against broad permissions like Bash(git:*).
  • Remote Code Execution (SAFE): The skill does not contain executable code, download scripts from external sources, or execute remote commands.
  • Privilege Escalation (SAFE): No attempts to acquire elevated permissions (sudo, chmod) detected. The skill explicitly encourages least-privilege principles for Bash tool usage.
  • Indirect Prompt Injection (SAFE): While the skill processes user input to generate new files, it includes mandatory clarification steps and scope validation (Steps 1 and 2), which serve as effective mitigations against processing malicious data blindly.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 05:58 PM