planner

Pass

Audited by Gen Agent Trust Hub on Jun 25, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns or security vulnerabilities were identified in the skill. The logic is consistent with its stated purpose of assisting in DigitalOcean deployments.
  • [COMMAND_EXECUTION]: The skill utilizes official command-line tools including doctl, gh, docker, and git. These are standard tools for managing DigitalOcean infrastructure and GitHub repositories. The commands are used appropriately for building, deploying, and validating services.
  • [EXTERNAL_DOWNLOADS]: The skill references official DigitalOcean and GitHub infrastructure, such as the digitalocean/app_action/deploy GitHub Action. These are well-known and trusted sources for deployment automation.
  • [CREDENTIALS_UNSAFE]: Static analysis flagged a database connection string in templates/tier-2-database/05-cloud-config.md.template. Upon manual review, this is a false positive. The string uses placeholders like {{APP_DB_USER}} and <PASSWORD>, containing no hardcoded secrets.
  • [SAFE]: The skill implements a secure approach to session state by directing the agent to store captured credentials (IDs, passwords, URLs) in a local .env.secrets file. The instructions explicitly mention that this file is gitignored, ensuring that sensitive data is not accidentally committed to version control while still allowing for session resumability.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 25, 2026, 01:56 PM
Security Audit — agent-trust-hub — planner