instantly-campaign-qc

Pass

Audited by Gen Agent Trust Hub on Mar 5, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection (Category 8).
  • Ingestion points: Lead data is ingested from the Instantly API using the list_leads tool as described in Step 4.
  • Boundary markers: No boundary markers or 'ignore' instructions are used to separate ingested lead data from the agent's instructions.
  • Capability inventory: Capabilities are limited to campaign and lead retrieval via MCP tools; no evidence of arbitrary command execution or unauthorized network access was found.
  • Sanitization: No sanitization or filtering of lead payload values is performed prior to the analysis and reporting phase.
  • [EXTERNAL_DOWNLOADS]: The skill utilizes ToolSearch to identify and load specific MCP tools, such as mcp__instantly-grundwerk__get_campaign and mcp__instantly-birdsview__get_campaign. These resources represent legitimate vendor-provided functionality for interacting with the Instantly platform.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 5, 2026, 01:19 PM