runpod
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill utilizes
npx @useclawlink/clito download and execute the ClawLink command-line interface, which is required for authenticating and interacting with the RunPod API. - [COMMAND_EXECUTION]: The instructions direct the agent to execute shell commands using the ClawLink CLI to perform actions such as listing GPU types, creating clusters, and managing secrets on the RunPod platform.
- [INDIRECT_PROMPT_INJECTION]: The skill acts as a bridge to external data from RunPod (e.g., pod details, cluster configurations) and possesses write capabilities that could be influenced by external data content.
- Ingestion points: Resource data and account details are fetched from RunPod via the CLI in
SKILL.md. - Boundary markers: The skill explicitly instructs the agent to "Read first; confirm with the user before any write."
- Capability inventory: The toolset includes the ability to create secrets, clusters, serverless endpoints, and update user settings through the CLI wrapper.
- Sanitization: Input is passed to the CLI using JSON strings in shell commands.
Audit Metadata