outreach-research

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill performs legitimate business research tasks using restricted tools and does not exhibit any malicious patterns.
  • [INDIRECT_PROMPT_INJECTION]: The skill possesses a data ingestion surface that processes user input and external connector outputs.
  • Ingestion points: User-supplied Ideal Customer Profile (ICP) and company domains in SKILL.md.
  • Boundary markers: Explicit instructions are provided to normalize hostnames, reject malformed values, and obtain user confirmation of the scope before execution.
  • Capability inventory: Functionality is confined to list_connectors and research_domain MCP tools; no file-system write access or shell execution capabilities are present.
  • Sanitization: Instructions specify input parsing, normalization, and deduplication logic to maintain data integrity.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 06:27 PM
Security Audit — agent-trust-hub — outreach-research