The Agent Skills Directory

[COMMAND_EXECUTION]: The skill enables the execution of arbitrary commands on remote targets through methods such as terminal:write and exec-winrm. This functionality is central to the skill's purpose as a remote management tool.
[EXTERNAL_DOWNLOADS]: The documentation contains a link to download a prebuilt binary of the websocat utility from the author's GitHub repository (https://raw.githubusercontent.com/DrOlu/agent-tools/main/websocat.exe).
[REMOTE_CODE_EXECUTION]: The agent:startTask RPC methods allow an AI agent to autonomously plan and execute multi-step operations on remote systems.
[PROMPT_INJECTION]: The skill processes userInput from WebSocket clients to drive agent tasks, representing an attack surface for indirect prompt injection. This risk is addressed by the RTerm backend's Command Policy engine and an AGT policy engine for governance.
[CREDENTIALS_UNSAFE]: The skill manages credentials via a secrets vault (observability:secretsSet). Documentation and example scripts use the placeholder RTERM_SECRETS_MASTER_KEY=pw to demonstrate master key configuration.

rterm-gateway