The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill documentation includes a link to an official documentation site at https://claw.227727.xyz. This domain has been identified as malicious and blacklisted by automated reputation scanners.
[REMOTE_CODE_EXECUTION]: The installation instructions for the Hermes Agent integration use a high-risk pattern (curl | bash) to download and execute a script from a remote GitHub repository. This method provides no opportunity for the user or system to verify the integrity or safety of the code before execution.
[COMMAND_EXECUTION]: The skill implements a 'Web CLI Terminal' using xterm.js, which provides a full terminal interface for the Hermes CLI. This allows for the execution of arbitrary commands on the host system where the gateway is deployed, creating a significant attack surface if the interface is exposed or compromised.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection due to the way it processes external data.
Ingestion points: Untrusted data enters the agent context through session metadata, chat message streams, and memory document updates (memoryApi).
Boundary markers: There are no boundary markers or instructions to ignore embedded commands within the processed data.
Capability inventory: The skill possesses powerful capabilities including terminal execution, file system management, and scheduled task triggering.
Sanitization: No sanitization or validation logic is present to filter malicious instructions from user-provided or external content before it is processed by the agent.

openclaw-admin-vue