skills/mukul975/anthropic-cybersecurity-skills/implementing-rapid7-insightvm-for-scanning/Gen Agent Trust Hub
implementing-rapid7-insightvm-for-scanning
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or unauthorized data exfiltration were detected. The scripts and documentation align with the stated purpose of vulnerability management and administrative automation.
- [INDIRECT_PROMPT_INJECTION]: The skill includes scripts that ingest vulnerability data from the Rapid7 InsightVM API, which represents a theoretical attack surface for indirect prompt injection if the source console were compromised.
- Ingestion points:
scripts/agent.py(get_site_vulnerabilities) andscripts/process.py(get_asset_vulnerabilities) fetch vulnerability titles and descriptions from the external API. - Boundary markers: Absent in the script output logic.
- Capability inventory: The scripts possess network access capabilities to the InsightVM console and file-system write access for generating reports (CSV/JSON).
- Sanitization: The data retrieved from the API is processed and reported without specific sanitization or escaping filters.
Audit Metadata