language-framework-specialist

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE] (SAFE): The analyzed files consist entirely of educational and instructional documentation. They provide safe coding patterns, architectural advice, and style enforcement for software development.
  • [PROMPT_INJECTION] (SAFE): No malicious instructions or bypass attempts were found. The 'specialized' persona files (e.g., specialized_Python_python-pro.md) are standard system prompt extensions designed to focus the agent's expertise on specific technical domains without subverting its core safety constraints.
  • [EXTERNAL_DOWNLOADS] (SAFE): The documentation mentions installation commands for standard, reputable software packages (e.g., pip install aiohttp, npm install prisma). All URLs provided are for official documentation or established GitHub repositories (e.g., laravel.com, github.com/psf/black).
  • [CREDENTIALS_UNSAFE] (SAFE): The guides emphasize security best practices, such as using environment variables for sensitive configuration (e.g., process.env.PRIVATE_KEY in the Hardhat guide or env('DATABASE_URL') in the Prisma guide) and avoiding hardcoded secrets.
  • [DATA_EXFILTRATION] (SAFE): No logic for unauthorized data collection or exfiltration was detected. Scraping examples (BeautifulSoup, Cheerio) are instructional and focus on ethical practices and rate limiting.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:36 PM