etl-connector-strategy
Pass
Audited by Gen Agent Trust Hub on Sep 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow that ingests external, untrusted data to validate demand for integration connectors.\n
- Ingestion points: The 'Interview' section and '1. Demand-validation gate' in
SKILL.mdprompt for user-provided data such as support tickets, churn notes, and API logs.\n - Boundary markers: Absent. The instructions do not provide delimiters or specific guidelines for the agent to distinguish between valid data and potentially embedded instructions.\n
- Capability inventory: The skill is documentation-centric and does not contain any scripts, subprocess calls, network requests, or file-writing operations across its 5 files.\n
- Sanitization: Absent.\n
- Note: The risk is negligible as the skill lacks any dangerous capabilities that could be targeted by an injection attack.\n- [SAFE]: The skill's behavior is consistent with its stated purpose of strategic consulting. All external service references (e.g., Fivetran, Airbyte) and internal vendor skill links are legitimate. No indicators of obfuscation, privilege escalation, or persistence mechanisms were found.
Audit Metadata