configure-notifications

Pass

Audited by Gen Agent Trust Hub on Jun 24, 2026

Risk Level: SAFECOMMAND_EXECUTIONCREDENTIALS_UNSAFE
Full Analysis
  • [CREDENTIALS_UNSAFE]: The skill is designed to handle and store sensitive credentials, including Telegram bot tokens and webhook URLs for Discord and Slack. These are saved to ~/.gemini/antigravity/.omc-config.json to enable notification functionality. This aligns with the skill's primary purpose for configuration management.
  • [COMMAND_EXECUTION]: The skill includes a 'Generic CLI Command' feature allowing users to configure custom shell commands for notifications. The instructions include security validations such as disallowing spaces and shell metacharacters in the executable name to mitigate accidental command injection.
  • [EXTERNAL_DOWNLOADS]: The skill uses curl to interact with official API endpoints for Telegram, Discord, and Slack to verify configurations. These are well-known services and the usage is limited to the skill's functional requirements.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests user input via the AskUserQuestion tool. It uses jq with argument passing (--arg) to safely assemble configuration files, preventing potential shell injection or JSON structure corruption from untrusted input.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 24, 2026, 12:00 AM
Security Audit — agent-trust-hub — configure-notifications