
Orchestrate
This is pattern documentation for running 20, 40, 60 agents in parallel without creating chaos. The strategies are mined from 597 real dispatches and map work type to orchestration pattern: research swarms for knowledge gathering (all background, each agent writes one doc), epic parallel builds for independent features (wave dispatch by subsystem), sequential pipelines for dependent work, parallel sweeps for the same transform across modules. The core insight is that most orchestration failures come from letting agents step on each other's files or dispatching without enough context for true parallelism. If you're scaling past 3 agents or thinking about fan-out patterns, this gives you the dispatch cadence, prompt structure, and background/foreground rules that actually work in production.
- skills.sh installs
- 669
- GitHub stars
- 36
- last updated
- Aug 21, 2026
Install
$ npx -y skills add hyperb1iss/hyperskills --skill orchestrate --agent claude-codeInstalls into .claude/skills of the current project.
Files
Multi-Agent Orchestration
The coordinator owns integration and the final claim of completion.
The user's instructions take precedence over this skill; delegation, tools, and external actions remain subject to current host instructions and authorization, and this skill grants a worker no authority the coordinator lacks.
Check Capabilities Before Dispatch
Inspect the actual tools and current schemas instead of inferring capabilities from a product name. An external CLI reviewer is still delegation, not an escape hatch.
Choose the Shape by Dependency
Sponsors
Details
- Repository
- github.com/hyperb1iss/hyperskills
- Categories
- DevOps & CI/CD, Git & Pull Requests, Code Review & Quality, AI & Agent Building, Data Science & ML, Debugging, Automation & Workflows
- First seen
- Apr 16, 2026