Skip to main content
skill CX Codex

planner · jdrhyne — Claude Code skill: what it does, how to install

planner structures complex projects into sprints and atomic tasks with acceptance criteria, preparing ready-to-run plans for subagents.

MIT ★ 239 updated 2026-08-30

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add jdrhyne/agent-skills --skill planner -a claude-code
Open on GitHub

What planner does

The skill guides project planning through five sequential phases: clarifying requirements with up to five targeted questions, inspecting the codebase, assembling detailed sprints, running a subagent review, and formatting the output. Each generated task includes exact file paths, estimated complexity on a 1–10 scale, testable acceptance criteria, and rollback instructions.

Every sprint is designed to produce a runnable increment with its own demo checklist. The resulting Markdown format is structured directly for downstream execution by workers like parallel-task or task-orchestrator.

The skill strictly separates planning from code edits: it does not touch implementation files and will not write the plan to disk unless an output artifact is requested.

What it helps with

  • break down complex features into runnable sprints with explicit checklists
  • generate orchestrator-ready Markdown plans for parallel execution
  • map out file paths, dependencies, and validation steps for atomic commits
  • define testing strategies, potential failure risks, and rollback steps before coding

Example prompts for the agent

  • Create a detailed plan to refactor the billing service and save it to plan.md
  • Break down adding OAuth login into atomic tasks with exact file locations
  • Plan the database migration into sequential sprints ready for parallel workers

How to install planner

1Skills CLI (Vercel)

npx skills add jdrhyne/agent-skills --skill planner -a claude-code

2Manually: put the files in place

~/.claude/skills/planner/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/planner/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about planner

How to install planner in Claude Code?

Run npx skills add jdrhyne/agent-skills --skill planner -a claude-code in your terminal.

What does the planner skill do?

It researches the codebase, clarifies constraints, and generates an actionable sprint-based plan formatted for autonomous subagents.

Will planner automatically start implementing tasks?

No, safety constraints keep the skill focused purely on planning until you explicitly ask to proceed with execution.

Source text

SKILL.md · MIT
---
name: planner
description: Create structured plans for multi-task projects that can be used by the task-orchestrator skill. Use when breaking down complex work into parallel and sequential tasks with dependencies.
metadata: {"clawdbot":{"emoji":"📋"}}
permissions:
  - file_write: "May save the completed plan to a file when the user asks for a persisted artifact."
---

# Planner

Create structured, orchestrator-ready plans for multi-task projects.

**Source**: Adapted from am-will's codex-skills workflow patterns
**Pairs with**: task-orchestrator skill for follow-on implementation

---

## Quick Start

Follow this process:

1. **Phase 0**: Clarify requirements (ask up to 5 targeted questions)
2. **Phase 1**: Research & understand the codebase
3. **Phase 2**: Create detailed plan with sprints, tasks, acceptance criteria
4. **Phase 3**: Subagent review of the plan
5. **Phase 4**: Return the plan in markdown and save a file only if the user asked for a persisted artifact

---

## Key Principles

### Task Atomicity
Each task must be:
- **Atomic and committable** — small, independent pieces of work
- **Specific and actionable** — not vague
- **Testable** — include tests or validation method
- **Located** — include file paths and code locations

### Bad vs Good Task Breakdown

❌ Bad: "Implement third-party sign-in"

✓ Good:
- "Add sign-in config to environment variables"
- "Install and configure the required authentication package"
- "Create sign-in callback route handler in src/routes/auth.ts"
- "Add the sign-in button to the login UI"

### Sprint Structure

Each sprint must:
- Result in a **demoable, runnable, testable** increment
- Build on prior sprint work
- Include clear demo/verification checklist

---

## Plan Template

```markdown
# Plan: [Task Name]

**Generated**: [Date]
**Estimated Complexity**: [Low/Medium/High]

## Overview
[Brief summary of what needs to be done and the general approach]

## Prerequisites
- [Dependencies or requirements that must be met first]
- [Tools, libraries, or access needed]
- [Tooling limitations, e.g., browser relay/CDP restrictions]

## Sprint 1: [Sprint Name]
**Goal**: [What this sprint accomplishes]
**Demo/Validation**:
- [How to run/demo this sprint's output]
- [What to verify]

### Task 1.1: [Task Name]
- **Location**: [File paths or components involved]
- **Description**: [What needs to be done]
- **Perceived Complexity**: [1-10]
- **Dependencies**: [Any previous tasks this depends on]
- **Acceptance Criteria**:
  - [Specific, testable criteria]
- **Validation**:
  - [Test(s) or alternate validation steps]

### Task 1.2: [Task Name]
[...]

## Sprint 2: [Sprint Name]
[...]

## Testing Strategy
- [How to test the implementation]
- [What to verify at each sprint]

## Potential Risks
- [Things that could go wrong]
- [Mitigation strategies]

## Rollback Plan
- [How to undo changes if needed]
```

---

## Hand-off

Once plan is ready, hand off to the **parallel-task worker**:

```
Please run parallel-task.md against my-plan.md
```

Or invoke directly:
> "Run all unblocked tasks in plan.md using parallel subagents. Keep looping until all tasks are complete."

## Safety Boundaries

- Do not save a plan file unless the user asked for one or the surrounding workflow explicitly needs a persisted artifact.
- Do not assign overlapping write scopes to parallel tasks without calling out the conflict.
- Do not invent dependencies, validation steps, or completion status that the repo context does not support.
- Do not turn a planning request into implementation work unless the user explicitly asks to move from planning to implementation.

---

## Related

- [`parallel-task`](../parallel-task/) — Parallel task worker that executes the plan this skill produces

Adapted from am-will's codex-skills planner/parallel-task prompts.

© jdrhyne · MIT · jdrhyne/agent-skills

Similar skills

All skills for Agent setup: memory, context, orchestration →

planner by other authors comes first.

From the same set27

The whole set →
clawdbot-release-check clawdbot-release-check watches GitHub for new OpenClaw releases and pings you once per version. Can run on a cron schedule and send the notice to Telegram. CX Codex +1 ★ 239repo stars Zendesk Work with Zendesk tickets and users through the API, or export queue data on request. Configure the required environment variables before connecting. CX Codex +1 ★ 239repo stars knowledge-graph The knowledge-graph skill keeps durable facts about people, companies, and projects in a workspace graph. Use it to carry useful context between sessions and look it up without reopening long transcripts. CX Codex +1 ★ 239repo stars self-improving-agent self-improving-agent keeps a small journal in .learnings/ with failures, user corrections and missing-capability requests, then reads it back before big jobs. CX Codex +1 ★ 239repo stars parallel-task Parses a markdown plan into sprint tasks and hands unblocked ones to parallel subagents. Fires on the explicit /parallel-task command. CX Codex +1 ★ 239repo stars elegant-reports elegant-reports turns markdown into a clean PDF report — executive brief, analysis write-up or slide-style deck — with bundled templates and themes. CX Codex +1 ★ 239repo stars gsc gsc enables read-only Google Search Console analysis in Claude Code across properties, search metrics, URL inspection, and sitemaps. CX Codex +1 ★ 239repo stars senior-engineering senior-engineering is a Claude Code skill for non-trivial tasks: it defines done first, then keeps the project runnable and checks a gate list before reporting completion. CX Codex +1 ★ 239repo stars

Project sponsors

Companies that keep this analytics open