Orchestrates parallel multi-agent workflows with delegation and persistence.
npx skills add https://github.com/MeroZemory/oh-my-droid --skill ultrawork-merozemory
Please help me install this Agent Skill. Skill: ultrawork Source: https://github.com/MeroZemory/oh-my-droid/tree/main/skills/ultrawork Command: npx skills add https://github.com/MeroZemory/oh-my-droid --skill ultrawork-merozemory
This skill keeps high-throughput Workflows moving by orchestrating multiple agents in parallel, aggressively delegating tasks, and enforcing persistence so no complex job stalls or needs constant manual supervision.
Ask for maximum performance mode to run tasks with aggressive parallel agent orchestration.
High-intent search queries and answers about installing and using this skill.
To run parallel agent workflows, you can enable maximum performance mode to launch independent agents simultaneously for every subtask. This ensures time-consuming operations progress together without stalling or requiring manual supervision.
Smart model routing in multi-agent orchestration immediately routes work to specialized tiers like LOW, MEDIUM, or HIGH. This matches expertise to complexity and conserves tokens while aggressively delegating tasks across the parallel droid teams.
To manage background operations, long-running commands are sent to background mode with enforced verification checklists. The system handles state cleanup by deleting ultrawork state files when the job is complete to avoid stale sessions.
Yes, parallel execution is specifically designed for Factory Droid multi-agent workflows. It targets simultaneous execution, aggressive delegation, and background operations to ensure no pending work remains during high-throughput tasks.
You should use aggressive delegation when building, testing, and reviewing a multi-component system. It enables persistent performance mode so every step checks off the verification list before completion, ensuring no complex job stalls.
Complex multi-agent jobs stall without manual supervision when persistence is not enforced. By orchestrating multiple agents in parallel and sending long-running commands to background mode, the workflow keeps moving automatically.
sigridjinethRoute AI tasks to lower-cost model tiers based on complexity.
shayne-snapAutomate end-to-end coding tasks with verification and architect approval.
suhwanOrchestrate Haiku and Sonnet agents for parallel, token-efficient task execution.
mazenyassergithubRoute tasks to specialist agents with parallel execution and background operations.
