task-orchestra
por tobisamaa
Productivity & TasksResumen
Coordinate multiple agents and tasks for complex workflows.
Descripción
Qué hace esta skill
Coordinates multiple agents and tasks for complex, multi-step workflows. Handles parallel execution, dependency resolution, state tracking, and error recovery across subagents. Supports sequential, pipeline, and supervisor orchestration patterns.
Handles dependency ordering and failure recovery automatically, removing the need to manually sequence or restart agents when one step fails.
Cuándo usarla
- Running a five-agent research project with parallel data collection
- Building a content pipeline from topic research to final publication
- Coordinating analysis, coding, and QA agents on a software feature
- Retrying or rerouting a failed subagent mid-workflow without restarting from scratch
- Resolving task dependencies before executing a multi-step data analysis
Flujo de trabajo de ejemplo
User asks: 'Orchestrate a research project with data collection, analysis, and report writing'
- 1
Break the project into subtasks: research, data collection, analysis, writing, QA review
- 2
Spawn a dedicated subagent for each subtask with scoped instructions
- 3
Run research and data collection agents in parallel
- 4
Pass collected outputs to the analysis agent once its dependencies resolve
- 5
Route analysis results to the writing agent and aggregate the final report
A completed research report produced by coordinated subagents, with progress tracked and partial failures handled automatically
Requisitos
- BRAVE_API_KEY environment variable
Instalación
clawhub install tobisamaa/task-orchestraEjecuta el comando de la CLI o pega el prompt en Claude Code, Cursor, Codex o el agente de programación que uses.