Implementation Plan & Subtask Generator

VerifiedSafe

Generates a PLAN.md and structured subtasks.json from a feature specification, breaking work into ordered, verifiable actions for agent execution.

Sby Skills Guide Bot
DevelopmentIntermediate
10/30/2026
Claude CodeCursorCodex
#planning#task-breakdown#spec-driven#project-management#implementation-plan

Recommended for

Automated summary

Generates a human-readable implementation plan (PLAN.md) and a structured subtask list (subtasks.json) from an existing Formic task's README.md specification.

Strengths

  • Produces two complementary artifacts: a human-facing PLAN.md and an agent-consumable subtasks.json
  • Defines an explicit JSON schema with IDs, statuses and timestamps, making progress tracking easy to automate
  • Guidance enforces single, verifiable subtasks and requires testing and quality-gate steps
  • Reads project development guidelines (kanban-development-guideline.md) when present to align with existing standards

Limitations

  • Never asks the agent to explore the actual source code, so file paths and function names can stay generic
  • Assumes a Formic task layout already exists ($TASK_DOCS_PATH with README.md) and does not fit other workflows
  • Overwrites PLAN.md and subtasks.json with no merge logic for an existing plan
When to use it

Use it when a feature specification is already written and you need to turn it into an implementation plan plus tracked subtasks for an agent to execute.

When not to use it

Avoid it during exploratory prototyping where scope is still undefined, or when no specification document has been written yet.

Security analysis

Safe
Quality score78/100

The skill only instructs the agent to read a README.md (and optional guideline file) and write two text files (PLAN.md, subtasks.json) under a task docs path. There is no shell execution, network access, credential handling, obfuscation, or destructive action. The templated $TASK_* variables are substituted by the host system, and the JSON schema shown is output data, not executable content. The content contains no instruction-injection attempts.

No concerns found

How this analysis is made

Examples

Plan a new auth feature
Read the spec in docs/tasks/auth-refresh/README.md and generate PLAN.md and subtasks.json with a phased implementation and testing steps.
Break down a task into subtasks
Create the implementation plan for task TASK-142 following our project guidelines, breaking it into 5-15 verifiable subtasks.

name: plan description: Generates implementation plan (PLAN.md) and structured subtasks (subtasks.json) for a Formic task.

Plan Skill - Generate Implementation Plan & Subtasks

You are a senior Technical Project Manager. Your task is to generate implementation planning documents based on an existing feature specification.

Task Title: $TASK_TITLE

Task ID: $TASK_ID

Task Docs Path: $TASK_DOCS_PATH


Instructions

Step 1: Read the Feature Specification

Read the README.md file at $TASK_DOCS_PATH/README.md to understand:

  • The task's goals and objectives
  • Key capabilities to implement
  • Non-goals (what to avoid)
  • Requirements to fulfill

Also read kanban-development-guideline.md in the project root if it exists for project-specific standards.


Step 2: Generate PLAN.md

Create a high-level implementation overview for human readers. Write to: $TASK_DOCS_PATH/PLAN.md

# [Task Title] - Implementation Plan

## Status
**PENDING** - Plan created, implementation to begin.

## Context
[1-2 sentences summarizing the task from README.md]

## Implementation Overview

### Phase 1: [First logical grouping]
[Brief description of what this phase accomplishes]

### Phase 2: [Second logical grouping]
[Brief description of what this phase accomplishes]

### Phase 3: [Third logical grouping if needed]
[Brief description of what this phase accomplishes]

## Key Milestones
- [Milestone 1]
- [Milestone 2]
- [Milestone 3]

## Success Criteria
- [Measurable outcome 1]
- [Measurable outcome 2]

Step 3: Generate subtasks.json

Create a structured subtask list that the agent will use as the source of truth during execution. Write to: $TASK_DOCS_PATH/subtasks.json

The subtasks.json file MUST follow this exact schema:

{
  "version": "1.0",
  "taskId": "$TASK_ID",
  "title": "$TASK_TITLE",
  "createdAt": "[ISO 8601 timestamp]",
  "updatedAt": "[ISO 8601 timestamp]",
  "subtasks": [
    {
      "id": "1",
      "content": "[Specific actionable task - be precise about files/functions]",
      "status": "pending"
    },
    {
      "id": "2",
      "content": "[Another specific actionable task]",
      "status": "pending"
    }
  ]
}

Subtask Guidelines:

  • Each subtask should be a single, verifiable action
  • Include specific file paths, function names, or component names when applicable
  • Order subtasks logically (dependencies first, then dependent tasks)
  • Include testing subtasks (e.g., "Write unit tests for auth service")
  • Include quality gate subtasks (e.g., "Verify all tests pass", "Run linter")
  • Aim for 5-15 subtasks depending on task complexity

Subtask Status Values:

  • pending - Not yet started (initial state for all subtasks)
  • in_progress - Currently being worked on
  • completed - Finished and verified

Guidelines

  • Be Specific: Use actual file names, function names, and paths from the codebase
  • Be Actionable: Each subtask should be a single, completable action
  • Be Realistic: Break down complex tasks into smaller steps
  • Follow Project Standards: Reference the project's development guidelines if available
  • Logical Ordering: Subtasks should follow a natural implementation order (backend before frontend, models before APIs, etc.)
  • Include Verification: Add subtasks for testing and quality gates

Output

Write both files:

  1. $TASK_DOCS_PATH/PLAN.md - Human-readable implementation overview
  2. $TASK_DOCS_PATH/subtasks.json - Structured subtask list for agent execution

Do not output anything else.

From the file to a first result

The download contains exactly the content displayed on this page. This copy may differ from the latest version in the source repository.

View the source and supporting files

This download contains SKILL.md only. Any scripts, references, templates and dependencies are not included. Get the required files from the source, preserving their directory structure, before continuing.

1. Download the file

Download SKILL.md

From the root of a test project, save the file as .skill-downloads/implementation-plan-subtask-generator-mvda880f/SKILL.md, or use the command below. This folder is for inspecting the skill before installation.

Command for macOS/Linux or WSL: requires a POSIX shell and curl. On Windows without WSL, use the download button and create the folder manually.

mkdir -p '.skill-downloads/implementation-plan-subtask-generator-mvda880f' && curl --fail --show-error --location 'https://skills-guides.com/api/skills/implementation-plan-subtask-generator-mvda880f/download' --output '.skill-downloads/implementation-plan-subtask-generator-mvda880f/SKILL.md.part' && mv '.skill-downloads/implementation-plan-subtask-generator-mvda880f/SKILL.md.part' '.skill-downloads/implementation-plan-subtask-generator-mvda880f/SKILL.md'

2. Check prerequisites

Read SKILL.md before using it. If it references scripts or external services, inspect them and prepare the required access in your test environment.

No recognized prerequisites section in this file. This does not mean it is self-contained: libraries, tools and access requirements may be mentioned elsewhere in the content or source repository.

The platforms on this page are catalog indications. The paths below follow tool documentation; they do not prove this skill has been run in those tools. No minimum tool version is confirmed for this listing.

3. Load the skill

After inspection, copy the complete folder into your project using one of the commands below, then open a new tool session. The command stops if a folder with the same name already exists: compare the files before replacing it.

Claude Code
test -f '.skill-downloads/implementation-plan-subtask-generator-mvda880f/SKILL.md' && test -s '.skill-downloads/implementation-plan-subtask-generator-mvda880f/SKILL.md' && mkdir -p '.claude/skills' && test ! -e '.claude/skills/plan' && cp -R '.skill-downloads/implementation-plan-subtask-generator-mvda880f' '.claude/skills/plan'

Cursor
test -f '.skill-downloads/implementation-plan-subtask-generator-mvda880f/SKILL.md' && test -s '.skill-downloads/implementation-plan-subtask-generator-mvda880f/SKILL.md' && mkdir -p '.cursor/skills' && test ! -e '.cursor/skills/plan' && cp -R '.skill-downloads/implementation-plan-subtask-generator-mvda880f' '.cursor/skills/plan'

Codex
test -f '.skill-downloads/implementation-plan-subtask-generator-mvda880f/SKILL.md' && test -s '.skill-downloads/implementation-plan-subtask-generator-mvda880f/SKILL.md' && mkdir -p '.agents/skills' && test ! -e '.agents/skills/plan' && cp -R '.skill-downloads/implementation-plan-subtask-generator-mvda880f' '.agents/skills/plan'

OpenSkills

OpenSkills must already be installed. The local install/read procedure is checked with OpenSkills 1.5.0, which requires Node.js ≥ 20.6.0. Catalog slugs are not resolved automatically.

Official OpenSkills installation
openskills install './.skill-downloads/implementation-plan-subtask-generator-mvda880f'

openskills read 'implementation-plan-subtask-generator-mvda880f'

Reading should display the installed file content and its base directory. It does not run the skill or install its dependencies.

For an agent that reads AGENTS.md, then sync the index. This command changes AGENTS.md: review the diff before using it.

openskills sync

Manual method or another tool: attach SKILL.md to a conversation, or open it in an agent that can read files. Include any required supporting resources. No universal folder is recognized by every tool.

4. Try a first task

Use a short, synthetic input: an API endpoint, a small document or an interface brief. In a new conversation, attach the file or give the agent its path, then adapt this request:

Read .skill-downloads/implementation-plan-subtask-generator-mvda880f/SKILL.md and apply its instructions to this example: [describe a test input]. Identify any missing files, dependencies or access. Produce the result requested by the skill without inventing data or running unreviewed scripts.

Expected result: output matching the skill instructions, or a precise list of missing prerequisites. A read confirmation alone does not validate a first use. If an external resource is unavailable, that path remains untested.

Related skills