task_planner

Breaks large tasks into bounded slices dispatched to skills and subagents.

1|Updated Jun 17, 2026
One-click install
npx skills add https://github.com/Tcuzzo/HydraAgent_public --skill task-planner-tcuzzo
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: task_planner
Source: https://github.com/Tcuzzo/HydraAgent_public/tree/main/skills/task_planner
Command: npx skills add https://github.com/Tcuzzo/HydraAgent_public --skill task-planner-tcuzzo

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Large tasks overwhelm a single agent context, causing drift and incomplete work. This Skill decomposes big goals into bounded, verifiable slices that are dispatched to the right skills and subagents within the Hydra runtime. ## Core Features & Use Cases - Working-Memory Bundles: Builds compact context bundles so each slice carries only the information it needs. - Bounded Delegate Slices: Splits large tasks into scoped units routed through skill_list, skill_show, and skill_route. - Reusable Skill Node Contracts: Defines repeatable contracts for how slices are dispatched and verified. - Use Case: Given a request to refactor a large module, the planner enumerates the relevant skills, spawns subagents for each bounded slice, and verifies each result before marking the task complete. ## Quick Start Ask the agent to plan and break down a large coding task into skill-dispatched slices using the task planner.

Frequently Asked Questions about task_planner

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

FAQPage Schema
How do I break a large task into smaller agent tasks?▼

Use the task planner to decompose a large goal into bounded slices, each with its own working-memory bundle. Each slice is routed to an appropriate skill or subagent and verified before the overall task is marked complete.

How does task planning work with skill routing in Hydra?▼

The planner inspects available skills via skill_list and skill_show, then uses skill_route to dispatch each bounded slice to the best-matching skill. Results are verified by the planner after each delegation.

Can the task planner spawn subagents for parallel work?▼

Yes, the planner uses spawn_subagent to delegate bounded slices to subagents. Each subagent receives a scoped working-memory bundle so it operates only on its assigned slice.

What tools does the task planner have access to?▼

It can read files, list directories, grep, list and inspect skills, route to skills, spawn subagents, and run shell commands. These tools support planning, dispatch, and verification of delegated slices.

When should I not use a planning skill for a task?▼

Small, single-step tasks do not benefit from decomposition and add unnecessary overhead. Use the planner only when a goal spans multiple files, skills, or verification stages.