Skill Detail
askesis-orchestrator
Building orchestration layers that coordinate agents, tools, routing, and memory without turning the system into chaos.
What problem it solves
Multi-agent systems become fragile when responsibility is unclear and every component tries to do everything. askesis-orchestrator solves the problem of coordination by giving the system a clear routing layer for delegation, memory, and tool use.
How it works
- •Accept a high-level task and classify what kind of work it actually is.
- •Route the task to the right specialist, tool, or workflow instead of treating all work the same.
- •Preserve memory and state so the system can continue instead of resetting every turn.
- •Keep handoffs explicit so specialized agents know their scope and output requirements.
- •Recompose specialist outputs into a coherent response or execution path.
Use case from logs
From coding assistant to orchestration agent
Context: A single assistant model stopped being enough once the work expanded into memory, delegation, and multi-agent execution across sessions.
What happened: The system evolved into an orchestration layer with role boundaries, explicit chains of responsibility, job orchestration, and memory-backed continuity. Agents were treated as specialized teammates rather than one undifferentiated tool.
Outcome: That shift enabled autonomous coordination, cleaner delegation, and a more robust collaborative system that could persist and evolve over time.
Source
Pause in Collaboration — content/blog/pause-in-collaboration.md
Echo became an orchestration agent with their own memory system... Together we built... job orchestration systems — autonomous execution infrastructure that let Echo and the other agents I work with operate independently.
GitHub
Code examples for this skill will link out to GitHub once the public repo is ready.