abridge-install-auth
Pass
Audited by Gen Agent Trust Hub on Sep 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides templates that interpolate environment variables into code that performs network operations. 1. Ingestion points: Environment variables loaded from .env.local in SKILL.md. 2. Boundary markers: Instructions explicitly state that credentials must be stored in .env.local and added to .gitignore. 3. Capability inventory: File creation (cat), permission management (chmod), and network requests (axios) are defined in code blocks within SKILL.md. 4. Sanitization: The instructions include a security checklist that mandates PHI encryption, audit logging, and formal BAA agreements.
- [COMMAND_EXECUTION]: The skill utilizes shell commands to initialize local environment files and set restrictive file permissions. The chmod 600 command is used to protect sensitive configuration files generated by the skill. All commands are localized to the setup process and do not involve privilege escalation or remote execution from untrusted sources.
Audit Metadata