Handle-Large-Tasks

Split large plans into subagents with tests and coherence checks.

149|10|Updated Oct 30, 2025
One-click install
npx skills add https://github.com/tilework-tech/nori-profiles --skill handle-large-tasks
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: Handle-Large-Tasks
Source: https://github.com/tilework-tech/nori-profiles/tree/main/src/installer/features/profiles/config/_mixins/_swe/skills/handle-large-tasks
Command: npx skills add https://github.com/tilework-tech/nori-profiles --skill handle-large-tasks

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you manage large, complex development tasks by breaking them into smaller, manageable chunks, effectively overcoming AI context window limitations and ensuring coherent, well-tested outcomes.

Core Features & Use Cases

  • Subagent Orchestration: Guides the creation and management of multiple subagents, each with a precise, limited task.
  • Context Management: Optimizes context window usage by delegating specific parts of a large task to individual subagents.
  • Test-Driven Subagents: Requires writing tests for each subagent's expected behavior, ensuring quality and correctness.
  • Use Case: When tasked with building a new microservice, use this skill to break it down into components (e.g., API, database, authentication), assigning each to a subagent with clear instructions and tests.

Quick Start

I have a large task to implement a new user dashboard. Help me break it down using subagents.

Frequently Asked Questions about Handle-Large-Tasks

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

FAQPage Schema
How do I break down a large software project to manage context limits?

Breaking down large projects involves splitting them into smaller, independent tasks assigned to subagents, each with clear scope and tests. This approach overcomes context window limitations by delegating specific components—such as API layers, database schemas, or authentication modules—to focused agents that work in parallel, then integrating results.

What's the best way to organize multiple subagents for a complex development task?

Organize subagents by functional component: define precise responsibilities for each (e.g., API handler, data layer, auth service), write tests upfront to specify expected behavior, then execute each subagent independently with the Task tool. This test-driven approach ensures coherence when components integrate.

Can I use subagents to handle tasks that exceed a single AI context window?

Yes. Subagent orchestration is designed specifically for this. By splitting a task into discrete, bounded units with clear inputs and tests, each subagent operates within its own context, making it possible to tackle projects—like building microservices or large features—that would otherwise exhaust a single context window.

How do I verify that work from multiple subagents integrates correctly?

After each subagent completes its task, review and iterate on outputs against the original test suite. Run integrated tests across all components to confirm coherence, then use restarts to reuse prior plans if adjustments are needed. This iterative feedback loop ensures the final result is unified and correct.

What's required before assigning work to subagents?

Define clear, testable requirements for each subagent: write tests that specify expected behavior upfront, then document precise task boundaries and inputs. This test-driven approach ensures subagents know exactly what to build and makes it straightforward to verify their outputs integrate properly.

When should I split a project into subagents rather than handle it in one go?

Split projects when they span long horizons or require multi-component orchestration that exceeds your context window. Examples include microservice builds, multi-layer feature implementations, or any task requiring coordination across distinct, independently testable components.