Loading...
Loading...
Found 13 Skills
Use when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies
This skill should be used when dispatching autonomous development or review tasks from GitHub issues. Covers scanning for new issues with the 'autonomous' label, dispatching dev-new/dev-resume/review processes, dependency checking, retry counting, stale process detection, and concurrency limiting. Use when asked to "run the dispatcher", "scan for pending issues", "dispatch autonomous tasks", "check stale agents", or "set up the dispatch cron".
Dispatch and coordinate parallel agent execution. Manages concurrent task processing with result aggregation and error handling.
Interactive command center for managing multiple phases from one terminal
Use when starting a session, deciding which framework skill applies to the current task, or sequencing them across a feature. Maps the user's intent to one of the five framework skills (ai-driven-prd, init-claude-project, generate-dev-plan, declarative-design, execute-plan) and enforces the cross-skill operating behaviors. Triggers on "which skill should I use", "where do I start", "how do these skills fit together", "I have a PRD now what", "/using-agent-skills".
Use when a task has multiple independent subtasks that can be executed concurrently by separate agents. Triggers when decomposed work has 2+ subtasks with no data dependencies, when subtasks operate on different files or codebase sections, when serial execution time would be significantly longer than parallel, or when independent analyses or deliverables need concurrent generation.
Dispatch independent tasks to concurrent subagents. Use when two or more failures or tasks are independent, sharing no files or state, and can be worked simultaneously.
Use when delegated work must survive a worker, Code Execution interruption, or session boundary and resume from durable plan files.
Use when running the CTO flow-optimization loop — WIP scan, epic focus, triage unlabeled issues, and dispatch.
Check progress, advance workflow, or dispatch freeform intent — the unified GSD situational command
You MUST use this, and not the upstream superpowers subagent-driven-development skill, when executing implementation plans with independent tasks in the current session. Dispatched task and whole-branch reviews run the scrutinize-dispatch skill; the re-review is deliberately unrouted.
Select interaction modes, capability tiers, risk levels, dispatch shapes, review mechanisms, and limited fallback options for non-trivial Agent tasks with clear objectives, scopes, and authorization. Applicable to tasks that require choosing between direct handling by the main Agent, single-task agents, and limited parallel processing, or determining capability tiers and review intensity based on complexity and risk; this Skill also guides users to choose to create, preview, or skip when the Codex project lacks validated Agent configurations. Do not use for simple Q&A, read-only confirmation, or directly verifiable minor modifications.