data-scraper-agent

Pass

Audited by Gen Agent Trust Hub on Apr 14, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill uses environment variables (GEMINI_API_KEY, NOTION_TOKEN, NOTION_DATABASE_ID) to manage credentials, ensuring that sensitive keys are not hardcoded in the script logic.
  • [COMMAND_EXECUTION]: The skill includes a GitHub Actions workflow (.github/workflows/scraper.yml) that executes shell commands to install dependencies, run the scraper, and commit feedback history. These operations are restricted to the intended automation and follow standard development practices.
  • [PROMPT_INJECTION]: The skill architecture includes a surface for indirect prompt injection through its web scraping functionality.
  • Ingestion points: Data is ingested from external URLs via requests and BeautifulSoup within the scraper/sources/ directory.
  • Boundary markers: The LLM prompt template in ai/pipeline.py uses markdown headers (# Items, # User Context) to separate external data from system instructions.
  • Capability inventory: The agent can perform network requests, write to external Notion databases, and update its own repository state via Git.
  • Sanitization: External data is processed via json.dumps() before being interpolated into the prompt, which helps prevent basic structural injection.
  • [EXTERNAL_DOWNLOADS]: The skill documents the use of standard, well-known Python packages (e.g., requests, beautifulsoup4, notion-client) and handles browser dependencies through the official Playwright CLI.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 14, 2026, 12:19 AM
Security Audit — agent-trust-hub — data-scraper-agent