plan

Turns large objectives into living Epic roadmaps with capped executable tasks.

1|1|Updated Jul 23, 2026
One-click install
npx skills add https://github.com/nearmap/shipyard --skill plan-nearmap
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: plan
Source: https://github.com/nearmap/shipyard/tree/main/skills/plan
Command: npx skills add https://github.com/nearmap/shipyard --skill plan-nearmap

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Large objectives are hard to execute directly: teams lose track of dependencies, over-decompose far work, and let roadmaps go stale as shipping evidence arrives. This Skill converts a big objective or existing Epic into a maintained roadmap on your issue tracker (Jira or GitHub Projects), keeping only a capped set of PR-sized tasks active at once. ## Core Features & Use Cases - Roadmap construction: Works backwards from a North Star through capabilities and dependency order to near-term executable leaves, with premise validation and prior-work/duplicate checks before shaping. - Living Epic maintenance: Re-enters an existing Epic, reconciles leaves as active, valid, blocked, or superseded, and appends decision-log checkpoints so the roadmap adapts to shipped evidence. - Capped parallel execution: Enforces the configured plan.max_active_tasks limit, identifies the critical path and parallel-safe set, and hands off each active leaf to /sy:spec. - Use Case: You have an Epic to migrate a monolith to services. Run the plan skill against the Epic key; it debates the shape, finds seams and hidden serialization, creates only the near PR-sized tasks, and gives you /sy:spec kickoff commands for each. ## Quick Start Ask the AI to plan a roadmap for your large objective or existing Epic key, for example by invoking /sy:plan with the objective text or Epic identifier.

Frequently Asked Questions about plan

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

FAQPage Schema
How do I break a large objective into an executable roadmap?▼

Invoke the plan skill with the objective or an existing Epic key. It validates the premise, checks for prior or duplicate work, finds seams and dependency order, then creates only near-term PR-sized tasks under the Epic and hands each off to /sy:spec.

How many tasks can be active in the roadmap at once?▼

At most the resolved plan.max_active_tasks cap of leaves may be active in spec, ship, or review simultaneously. If the near horizon exceeds the cap, only the critical or uncertainty-reducing subset is created and the rest stays conceptual in the Epic.

Does the plan skill work with Jira and GitHub Projects?▼

Yes. All tracker writes go through the tracker skill, which supports Jira and GitHub Projects. Epics, child tasks, dependencies, and decision-log comments are created on the configured tracker after explicit user approval.

What happens when I re-run planning on an existing Epic?▼

Re-entry reconstructs the delta since the last Plan checkpoint: it reads retrospectives from completed items, classifies every non-terminal leaf as active, valid, blocked, or superseded, and appends a new decision-log entry rather than rewriting history.

When should I not decompose work into new tasks?▼

Far work stays conceptual until evidence justifies decomposition, and small adjacent low-risk fixes found mid-ship fold into the current PR as a recorded scope extension instead of earning their own leaf.