sub-agents-mcp

Define task-specific AI sub-agents in Markdown for any MCP-compatible tool. Operations teams can automate workflows by creating specialized sub-agents that integrate with existing tools and systems. This skill connects to MCP-compatible tools and workflows, enabling efficient task automation.

[{"step":"Identify the specific task to automate (e.g., log parsing, API polling, data validation).","action":"Use the prompt template to generate a sub-agent definition by replacing [TASK_NAME], [MCP_TOOL_NAME], [SPECIFIC_SCENARIO], and [CONSTRAINTS].","tip":"Start with a narrow scope (e.g., 'validate API response format') before expanding to complex workflows."},{"step":"Customize the sub-agent's input/output schema, error handling, and integration steps for your MCP-compatible tools (e.g., database_utils, file_system, or custom APIs).","action":"Modify the Markdown template to include your tool's specific parameters, authentication requirements, and output formats.","tip":"Reference existing MCP tool documentation to ensure compatibility with input/output formats."},{"step":"Deploy the sub-agent to your MCP server and test with sample inputs.","action":"Save the Markdown file to your MCP `sub-agents/` directory, register it with `mcp register`, and invoke it with `mcp invoke`.","tip":"Use the `--dry-run` flag for initial testing to avoid unintended side effects."},{"step":"Integrate the sub-agent into your automation pipeline (e.g., cron jobs, CI/CD workflows, or monitoring systems).","action":"Chain sub-agents together using MCP's workflow capabilities or trigger them via API calls from your existing tools.","tip":"Start with a single sub-agent in a non-critical workflow to validate performance before scaling."},{"step":"Monitor performance and refine the sub-agent based on logs and error reports.","action":"Check MCP server logs for execution details and adjust parameters or error handling as needed.","tip":"Set up alerts for sub-agent failures to proactively address issues."}]