recoup-internal-consulting-lead-intake
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local Python script 'integrations/gmail/_work/pull_threads.py' using a '--query' argument constructed from potentially untrusted external lead names or domains, which could lead to command injection if the agent does not properly validate the input string.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted content from external sources which could contain malicious instructions.
- Ingestion points: The workflow processes 'fresh inquiry email', 'first-call transcript', and 'Gmail threads' which are externally controlled (SKILL.md).
- Boundary markers: The skill provides defensive instructions requiring the agent to tag data as '[UNVERIFIED]' or '[MISSING]' and to cross-reference multiple primary sources before taking action.
- Capability inventory: The agent has the ability to create and modify files in the 'pipeline/' directory, execute local scripts, and update records in the Attio CRM.
- Sanitization: There are no specific programmatic sanitization requirements or delimiters defined for the untrusted content; security relies on the agent's adherence to the verification instructions.
- [DYNAMIC_EXECUTION]: The skill depends on local adapter scripts located in 'integrations/' to bridge with external services, creating a dependency on the integrity of the local file system and environment.
Audit Metadata