opencode-supabase-guide

Pass

Audited by Gen Agent Trust Hub on Jul 5, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill defines a structured and secure workflow for Supabase integration using authenticated vendor tools like get_publishable_keys and supabase_list_projects.
  • [SAFE]: It strictly enforces the 'Question Tool Rule', which mandates that the agent must use an interactive tool for all confirmations and selections (e.g., project selection, browser opening). This prevents autonomous or hidden actions.
  • [SAFE]: No malicious external dependencies or remote script execution patterns were detected. The skill explicitly avoids unnecessary package installation by stating the functionality is bundled.
  • [SAFE]: The process for handling 'Studio Prompts' for MCP setup relies on the user providing data that is then parsed and written to local configuration files (.opencode/opencode.json), which is a standard and expected behavior for this service.
  • [SAFE]: The authentication process requires the user to manually execute a command in their terminal (opencode mcp auth supabase) and complete OAuth in a browser, maintaining a high level of user control and transparency.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 5, 2026, 03:40 PM
Security Audit — agent-trust-hub — opencode-supabase-guide