close-plan-generator-skill
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources, which could potentially contain malicious instructions designed to influence the agent's behavior.
- Ingestion points: The skill ingests data from HubSpot deal notes (hubspot_get_deal), contact records, and Clari call intelligence summaries (clari_get_call_summary) to build its reports.
- Boundary markers: The instructions do not define specific delimiters or instructions to ignore embedded prompts within the retrieved CRM or call data.
- Capability inventory: The skill possesses the ability to create Gmail drafts (create_draft) and modify HubSpot deal information.
- Sanitization: Risk is significantly mitigated by 'Gate B' (brand check) and 'Gate D' (draft only), which ensures that any generated customer-facing text is reviewed by a human (AE) before being sent, and is checked for brand alignment using the check_my_copy tool.
Audit Metadata