use-task-gateway
Pass
Audited by Gen Agent Trust Hub on Oct 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is a set of integration instructions for a task gateway service provided by xgent-ai. It does not contain malicious code, hidden instructions, or unauthorized data access patterns. All operations described are legitimate within the context of the platform's service ecosystem.
- [INDIRECT_PROMPT_INJECTION]: The skill describes a system that processes task-related data, which creates a potential attack surface. However, the documentation includes specific mitigations to address this risk:
- Ingestion points: Task payloads and metadata fields submitted to the gateway's API as described in references/task-api.md.
- Boundary markers: The API utilizes structured JSON and the gateway is explicitly instructed not to interpret payload content as direct commands.
- Capability inventory: The gateway supports CLI execution, API integration, and SSH-based agent deployment for worker nodes as detailed in references/workers.md.
- Sanitization: Instructions explicitly advise against trusting absolute paths or shell strings from payloads, recommending the use of fixed command arrays and logical identifier mapping to restricted directories.
Audit Metadata