tam-building

Pass

Audited by Gen Agent Trust Hub on Sep 1, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [SAFE]: The skill utilizes the Cargo CDK framework to define models and workflows. It uses adopted connectors for Sales Navigator and Cargo's database, which are intended for its primary purpose of data sourcing and enrichment.
  • [EXTERNAL_DOWNLOADS]: The skill instructions and code reference external domains such as linkedin.com and ai-ark.com. These references are part of the core functionality for searching and enriching company records.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external LinkedIn searches to populate a database. While this creates an entry point for external content, the data is processed as structured fields within a controlled workflow.
  • Ingestion points: LinkedIn Sales Navigator search results defined in infra/models/salesnav-companies.ts.
  • Boundary markers: Data is processed as distinct string fields during the promotion workflow.
  • Capability inventory: Database upserts and business enrichment API calls in infra/plays/promote-to-accounts.ts.
  • Sanitization: Input validation using Zod and standard field mapping within the workflow.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 1, 2026, 02:51 PM
Security Audit — agent-trust-hub — tam-building