work
Pass
Audited by Gen Agent Trust Hub on Oct 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes multiple shell commands using the
hubspotCLI to interact with project management data. These operations, such ashubspot work records updateandhubspot auth login, are official vendor tools that run within the user's existing authentication context and follow standard access controls. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from HubSpot Work trackers, documents, and comments (ingestion points:
records search,docs get,comments list,tables get). While this creates a surface for indirect prompt injection, the skill maintains write capabilities (capability inventory:records create,records update,records complete,comments create) and mitigates risk through explicit instructions. The agent is required to use--dry-runfor modifications, present previews for user confirmation, and strictly separate factual data from AI inferences in its responses. No specific boundary markers or sanitization steps are defined for the external content.
Audit Metadata