labor-arbitration-agent
Pass
Audited by Gen Agent Trust Hub on Apr 18, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill package is composed entirely of Markdown documentation and instructions. No executable scripts (.py, .js, .sh), binaries, or configuration files that could facilitate code execution were found.
- [SAFE]: No network operations, data exfiltration patterns, or external downloads are present. All references to 'evidence collection' and 'bank records' refer to instructions for the user to manage their own documents within the context of a labor dispute.
- [SAFE]: There are no attempts at prompt injection or bypassing safety filters. The system prompt clearly defines the agent's role, boundaries, and professional ethics within the specified legal domain.
- [SAFE]: No obfuscation techniques, hidden URLs, or encoded content were detected across any of the 12 files.
- [SAFE]: Metadata and frontmatter are consistent with the skill's purpose, although the main SKILL.md file contains some remnant template text (e.g., 'Board Prep' workflow) which is a non-security related quality artifact.
Audit Metadata