skills/acedatacloud/skills/telegram/Gen Agent Trust Hub

telegram

Pass

Audited by Gen Agent Trust Hub on Jul 16, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill utilizes environment variables for all sensitive credentials (TELEGRAM_API_HASH, TELEGRAM_SESSION_STRING) and includes explicit instructions to the AI agent to never echo or log these secrets.
  • [SAFE]: A robust safety mechanism is implemented where all write or destructive actions (send, delete, edit, react) require a --confirm flag as the final argument. The script defaults to a dry-run mode, ensuring users can review the planned action before it is executed.
  • [SAFE]: The materialize_file function in the auxiliary script handles remote URL downloads securely by validating the protocol (restricting to HTTP/HTTPS) and using temporary file management with proper cleanup.
  • [SAFE]: The skill correctly manages indirect prompt injection risks common in messaging tools by requiring explicit user confirmation for any actions driven by the content of processed messages.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 16, 2026, 12:43 PM
Security Audit — agent-trust-hub — telegram