delegation

Dispatch implementation tasks across subagents with worktree isolation and TDD validation.

33|Updated Jan 5, 2026

Search Tags:#tdd #worktrees #subagents #delegation #orchestration #provenance #agent-teams

Ready to run

npx skills add https://github.com/lvlup-sw/exarchos --skill delegation-lvlup-sw

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: delegation
Source: https://github.com/lvlup-sw/exarchos/tree/main/skills-src/delegation
Command: npx skills add https://github.com/lvlup-sw/exarchos --skill delegation-lvlup-sw

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Dispatch implementation tasks across subagents or agent teams with strict worktree isolation and TDD-driven validation.

Core Features & Use Cases

Orchestrates Prepare, Dispatch, Monitor phases for feature work across multiple participants
Enforces per-task worktrees, deterministic prompts, and quality gates
Supports subagent and agent-teams modes for parallelism and collaboration
Use case: parallelizing a feature across 2 tasks with separate feature branches and CI checks

Quick Start

Provide an intuitive prompt to implement tasks? We can provide a high-level plan to delegate tasks for a feature across subagents and monitor progress in worktrees.

Frequently Asked Questions about delegation

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I coordinate multi-agent tasks across separate git worktrees?▼

To coordinate multi-agent tasks across worktrees, you dispatch implementation tasks to subagents or agent teams with strict per-task worktree isolation and TDD-driven validation. The system orchestrates preparation, dispatch, and monitoring phases while recording provenance for traceability.

Can I parallelize a feature across multiple subagents with separate branches?▼

Yes, you can parallelize a feature across multiple subagents by dispatching tasks into isolated worktrees with separate feature branches. The orchestration mode enforces deterministic prompts and quality gates, ensuring each parallel task runs independently with CI checks.

How does TDD-driven validation work with agent teams orchestration?▼

TDD-driven validation in agent teams orchestration works by enforcing quality gates during the monitoring phase. It ensures deterministic execution across dispatched tasks, validating that implementation meets test requirements before progress gating allows completion or triggers recovery paths.

What is the recovery path when a dispatched subagent task fails validation?▼

The recovery path for a failed subagent task uses fix-mode and operator gates. When validation fails during monitoring, the orchestration enforces deterministic execution and provides these recovery paths to handle errors without losing per-task context isolation or provenance traceability.

Do I need separate worktrees for each task in multi-agent orchestration?▼

Yes, separate worktrees are required for each task to maintain strict context isolation. The orchestration enforces per-task worktrees and deterministic prompts, ensuring parallel implementation across subagents or agent teams does not interfere and provenance remains traceable.

What's the best way to monitor progress across dispatched agent teams?▼

The best way to monitor progress across dispatched agent teams is through the orchestration's monitoring phase. It handles progress gating and records provenance, providing operator gates to oversee deterministic execution and ensure TDD-driven validation passes before task completion.