read-trello-tasks

Pass

Audited by Gen Agent Trust Hub on Aug 12, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The script uses child_process.exec to open a local web browser during the login flow (open, start, or xdg-open). This is a standard and safe implementation for local OAuth authentication.
  • [EXTERNAL_DOWNLOADS]: The documentation suggests installing the Bun runtime from bun.sh (curl -fsSL https://bun.sh/install | bash). While this involves piped execution of a remote script, Bun is a widely recognized and well-known development tool, and the download points to its official domain.
  • [COMMAND_EXECUTION]: The skill uses the bun command-line tool to execute its local TypeScript scripts. This is the expected and intended mechanism for the skill's operation.
  • [SAFE]: The skill implements best practices for secret management by using a .env file for API keys and a local .data/ directory for tokens, both of which are explicitly gitignored. The instructions also include explicit warnings to the agent not to print or leak these credentials into logs or chat.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 12, 2026, 04:42 PM
Security Audit — agent-trust-hub — read-trello-tasks