ai-agent/.roo/rules-enhanced-recursive-engineer/03-user-commands.md

1.5 KiB

/handoff - Initiate handoff process to a new agent instance (requires manual confirmation or high automation) /check-context - Report on current context utilization and estimated limit percentage (Use proactively if suspecting high usage) /search "query" - Perform a web search for the specified query using MCP tools /tools - List available tool groups and key tools within them /confidence - Report confidence assessment for the current task/solution /plan - Display the current high-level execution plan and progress step /debug - Enter focused debugging mode for the current issue (enhances logging and step-by-step reasoning) /test - Sequentially run tests relevant to the current implementation using execute_command /document - Generate documentation for the current implementation artifact /help - Display available commands and information /feedback "message" - Record user feedback for the internal learning system (appends to a specific learning feedback file) /explain - Provide detailed explanation of the current approach or reasoning /fastpath - Request expedited processing for a task identified as routine (uses pre-defined sequences) /delegate-to [agent-slug] "task" - Delegate a specific task to another agent (requires 'high' automation or explicit approval) /request-review "artifact-path" - Request review of a specific artifact (e.g., complex code module) via Conductor /escalate "issue-description" - Escalate an issue to a higher level (e.g., the agent that assigned the task)