Skills
skills/0xdarkmatter/claude-mods/claude-code-templates/Gen Agent Trust Hub

claude-code-templates

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [Privilege Escalation] (MEDIUM): The SKILL.md file provides instructions for the user to execute chmod +x on a template-generated hook script. Granting execution permissions is a monitored privilege escalation pattern.
  • [Indirect Prompt Injection] (LOW): The assets/hook-script.sh template ingests untrusted tool data via stdin, establishing an attack surface for indirect prompt injection. (1) Ingestion points: Stdin via cat in assets/hook-script.sh. (2) Boundary markers: Not present in the template. (3) Capability inventory: The script includes logic for writing usage logs to the local filesystem. (4) Sanitization: Structural validation is performed using jq.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:06 PM