integrate-agentkit-host
Pass
Audited by Gen Agent Trust Hub on Nov 4, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructions include commands to install external modules via
clawhubandhermes skills installfrom repositories hosted on GitHub and the ClawHub registry. - [COMMAND_EXECUTION]: The skill uses shell commands to install host-specific skills and execute the provided Python CLI tool (
scripts/tool_exec.py) using theuvpackage manager. - [DATA_EXFILTRATION]: The CLI tool transmits Scalekit client credentials (ID and Secret) from the environment to the user-specified
SCALEKIT_ENVIRONMENT_URLto perform an OAuth2 client credentials grant and obtain a bearer token for API access. - [INDIRECT_PROMPT_INJECTION]: The skill interacts with external APIs and tools, presenting an attack surface for indirect prompt injection through tool outputs.
- Ingestion points:
scripts/tool_exec.pyreceives and displays responses from the Scalekit tool execution and proxy request APIs. - Boundary markers: None are explicitly defined in the CLI tool output when displaying external data to the agent.
- Capability inventory: The CLI supports tool execution, proxied HTTP requests (including file uploads/downloads), and connection management.
- Sanitization: Input for tool execution is validated through JSON parsing in the CLI before being transmitted.
Audit Metadata