extract-skill-to-repository

Pass

Audited by Gen Agent Trust Hub on Aug 14, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructions require the agent to use Git commands to manage worktrees, commit changes, and potentially push to remote repositories. These operations are governed by repository-specific policies and user approval.
  • [PROMPT_INJECTION]: The skill processes untrusted data by reading and parsing the instructions and metadata of other agent skills during the extraction process. This creates an indirect prompt injection surface.
  • Ingestion points: Reads the source SKILL.md file and all contents within the .agents/skills/ directory.
  • Boundary markers: The skill uses YAML parsing for frontmatter but lacks specific markers to ignore embedded instructions in the skill body during the move process.
  • Capability inventory: Possesses filesystem access (read/write/delete) and Git version control capabilities (commit/push).
  • Sanitization: The procedure includes YAML validation, path resolution, mandatory hash verification, and repository-specific validation checks to mitigate content-based risks.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 14, 2026, 06:59 PM
Security Audit — agent-trust-hub — extract-skill-to-repository