customer-retention
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the official HubSpot CLI along with common shell utilities such as
jq,date, andpasteto query CRM records and manage task associations. These operations are consistent with the skill's primary purpose of CRM automation. - [INDIRECT_PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface due to its handling of customer records.
- Ingestion points: Contact data is ingested into the agent context via
hubspot objects searchcommands inSKILL.md. - Boundary markers: There are no explicit delimiters or instructions provided to isolate this retrieved data from the agent's internal logic.
- Capability inventory: The skill has the capability to write back to the CRM and modify object relationships using
hubspot objects createandhubspot associations create. - Sanitization: No sanitization or validation is applied to contact properties (e.g.,
firstname) before they are interpolated into new task subjects usingjq, which could allow malicious content within CRM fields to influence the agent's output.
Audit Metadata