static-residential-ip-assessor

Pass

Audited by Gen Agent Trust Hub on Mar 5, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it is designed to scrape and process content from external websites (e.g., AbuseIPDB, IPQS, and Scamalytics). Adversarial content on these third-party pages could potentially influence the agent's subsequent decisions or behavior. Ingestion points: automated browser queries to ipinfo.io, ip-api.com, abuseipdb.com, scamalytics.com, ipqualityscore.com, bgp.he.net, and rdap.arin.net. Boundary markers: absent; the instructions do not include delimiters or specific instructions to disregard embedded commands in the scraped data. Capability inventory: local file read/write access to ~/ip-assessments.md and utilization of a browser MCP tool for network navigation. Sanitization: absent; the agent extracts data directly from the rendered HTML of the external sources.
  • [EXTERNAL_DOWNLOADS]: The skill fetches IP intelligence data from several well-known and trusted technology services, including ipinfo.io, IPQualityScore, and Scamalytics.
  • [DATA_EXFILTRATION]: Accesses the local filesystem to read from and append to the file ~/ip-assessments.md for the purpose of maintaining a history of IP assessments across sessions.
  • [COMMAND_EXECUTION]: Instructs the agent to use browser automation capabilities (such as Playwright or Chrome DevTools MCP) to perform structured data collection from specific external URLs.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 5, 2026, 10:07 AM