gorilladesk-private-api

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external data from the GorillaDesk private API, creating an attack surface for indirect prompt injection.
  • Ingestion points: The agent retrieves job and customer data from endpoints such as ab2.gorilladesk.com/api/customers/{id}/jobs and ab2.gorilladesk.com/api/search/elastic.
  • Boundary markers: The instructions do not provide explicit delimiters or instructions to ignore embedded commands within the external data.
  • Capability inventory: The skill allows the agent to execute write operations, including PUT /api/jobs/{jobId}/status and POST /api/invoices.
  • Sanitization: The documentation mentions a programmatic guard (_refuse_if_sending()) in the referenced Python implementation to prevent accidental email triggers, but does not specify input sanitization for the LLM context.
  • [DATA_EXFILTRATION]: The skill performs network operations to ab2.gorilladesk.com and app.gorilladesk.com. These are non-whitelisted domains used for the skill's primary function of interacting with the CRM.
  • [EXTERNAL_DOWNLOADS]: The skill mentions the retrieval of a 23MB JavaScript source map from app.gorilladesk.com/static/js/main.<hash>.chunk.js.map to reverse-engineer private API payloads and logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 06:55 AM
Security Audit — agent-trust-hub — gorilladesk-private-api