env-secrets-manager
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The core functionality is implemented via a Python script (scripts/env_auditor.py) that performs local static analysis of repository content using predefined regular expression patterns to detect known credential formats for services like AWS, GitHub, and Slack.
- [SAFE]: The skill references and provides configuration guidance for established, well-known security tools and services including HashiCorp Vault, AWS Secrets Manager, Azure Key Vault, and Gitleaks.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses a data ingestion surface by reading arbitrary file content from a local repository for scanning. While this creates a potential vector for indirect injection, it is the intended and necessary behavior for an auditing tool and is handled via local static analysis rather than runtime execution. Ingestion points: Repository files read by scripts/env_auditor.py and grep-based shell scripts in reference files. Boundary markers: None implemented. Capability inventory: Local file system read access. Sanitization: None; the tool extracts and displays raw matches for human review.
Audit Metadata