skills/alanalvestech/hitank/trello/Gen Agent Trust Hub

trello

Pass

Audited by Gen Agent Trust Hub on May 9, 2026

Risk Level: SAFE
Full Analysis
  • [DATA_EXFILTRATION]: The skill implements secure credential management by storing Trello API keys and tokens in a local file (~/.config/trello/credentials.json). The save_token.rb script ensures the file is created with restricted permissions (0600), which prevents other users on the system from reading the secrets.
  • [PROMPT_INJECTION]: Because the skill processes and displays external content from Trello (such as card names, descriptions, and comments), there is an inherent risk of indirect prompt injection where the agent might interpret malicious content in a Trello card as instructions.
  • Ingestion points: External data enters the agent's context through Trello API responses processed in card.rb, cards.rb, comments.rb, and search.rb.
  • Boundary markers: The scripts print the Trello data to standard output without the use of explicit delimiters or boundary markers.
  • Capability inventory: The skill is authorized to make network requests to the Trello API and can perform destructive actions (like deleting cards) upon user confirmation.
  • Sanitization: Data retrieved from Trello is output directly without escaping or sanitization of potential command patterns.
Audit Metadata
Risk Level
SAFE
Analyzed
May 9, 2026, 01:11 AM
Security Audit — agent-trust-hub — trello