agency-digest-setup

Warn

Audited by Gen Agent Trust Hub on Feb 28, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONCREDENTIALS_UNSAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The setup process involves executing shell commands to create a Python virtual environment, install dependencies, and run the automation script.\n- [COMMAND_EXECUTION]: The skill configures a persistent task on macOS by creating a LaunchAgent plist file in ~/Library/LaunchAgents/ and loading it with launchctl. This allows the script to run automatically at 8 AM daily.\n- [CREDENTIALS_UNSAFE]: The workflow asks the user for Intelligems API keys and Slack Webhook URLs, which are then stored in local configuration files (brands.json and .env).\n- [EXTERNAL_DOWNLOADS]: The skill installs common Python packages (requests, python-dotenv) from the official PyPI registry during setup.\n- [PROMPT_INJECTION]: The skill processes data from the Intelligems API and includes it in Slack messages, which represents a potential surface for indirect prompt injection.\n
  • Ingestion points: Data is fetched from https://api.intelligems.io/v25-10-beta in agency_digest.py.\n
  • Boundary markers: No specific delimiters or instructions are used to separate API data from the agent's logic.\n
  • Capability inventory: The script has network access and is scheduled for recurring execution.\n
  • Sanitization: The script does not perform validation or escaping of the content received from the API before sending it to Slack.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 28, 2026, 08:15 PM