sequence-load
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted lead data (such as names, job titles, and company names) from the Apollo mixed people search API. This data is then displayed to the agent and processed for enrichment and sequence enrollment, creating a surface for potential prompt injection from external records.- Ingestion points: Contact data retrieved via mcp__claude_ai_Apollo_MCP__apollo_mixed_people_api_search in Step 4.- Boundary markers: Absent; there are no instructions to the agent to treat the retrieved contact data as strictly non-executable text or to ignore embedded instructions within that data.- Capability inventory: The skill can reveal personal emails, create contacts, and enroll them in email outreach campaigns using various Apollo MCP tools defined in SKILL.md.- Sanitization: No sanitization, validation, or escaping of the ingested lead data is specified before it is presented to the user or passed to subsequent tools.
Audit Metadata