agentleFS
supper-skillssupper-skills/develpment-skills/skills/writing-plan/SKILL.md

writing-plan

Skill2 starsChanged 8 months agoMITView on GitHub

Implementation plan creation skill for multi-step tasks. Use when there are specifications or requirements, before directly touching code. Create detailed, executable implementation plans. Activate when user mentions "create plan", "implementation plan", "write plan", "plan", "task breakdown", or starts complex multi-step development work with existing specifications.

1 other project uses a version of this.

Adopted by

Most recently taken up by

  • Other versions1 project uses a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 1

    1. rakaadi/agent-kit★ 1same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository4 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

12 rules. About 0 are common, 11 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈11

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Before the plan is completed, do NOT:

    line 19
  2. Distinctive:

    Create or modify project files (except plan documents)

    line 23
  3. Distinctive:

    Use a unified session state file to track progress:

    line 28
  4. Distinctive:

    After completing phase milestones → Update state file

    line 36
  5. Distinctive:

    Read specification document, understand requirements

    line 55
  6. Distinctive:

    Test files must be created before task starts

    line 171
  7. Distinctive:

    Completion Criteria: All tasks defined, each task includes complete test plan, user confirmed

    line 173
  8. Distinctive:

    If user approves, ask if execution should begin

    line 185
  9. Distinctive:

    Termination: After this phase completes, this skill ends.

    line 208
  10. Distinctive:

    When multiple subsystems are identified:

    line 214
  11. Distinctive:

    Update affected plan documents

    line 266
  12. Very specific:

    Read session state file design/.session-state.md

    line 273

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Writing Plan Skill

This skill is used to create comprehensive implementation plans for multi-step development tasks. Use when there are specifications or requirements, before directly touching any code.


Core Principles

Plan First

Before the plan is completed, do NOT:

  • Invoke any implementation skills
  • Write any code
  • Create or modify project files (except plan documents)
  • Take any implementation actions

Session State Management

Use a unified session state file to track progress:

  • Path: design/.session-state.md
  • Template: references/plan-template.md (shared)

When to Operate:

  • Skill activation → Create/update state file
  • After completing phase milestones → Update state file
  • At the start of each dialogue round → Read state file to restore context
  • After each phase completion → Update state file

Termination Condition

The only way to terminate this skill is to complete the plan document and notify user that execution is ready.


Workflow

Phase 1: Context and Scope Analysis

Goal: Understand specification document, identify subsystems.

Actions:

  1. Receive specification document path passed from brainstorming
  2. Read specification document, understand requirements
  3. Assess if multiple independent subsystems are involved

Subsystem Identification Criteria:

CharacteristicDescription
Independent Data ModelSubsystem has its own data structure
Independent InterfaceSubsystem provides independent API externally
Independent DeploymentSubsystem can be deployed and run independently
Clear BoundariesSubsystems communicate through explicit interfaces

Handling Strategy:

  • Single subsystem → Proceed directly to Phase 2
  • Multiple subsystems → Create outline file first, then write plans for each subsystem

Completion Criteria: Requirements understood, subsystem division identified


Phase 2: File Structure Planning

Goal: Plan files to be created or modified and their responsibilities.

Design Principles:

  • Single Responsibility: Each file should have a single clear responsibility
  • Context-friendly: Code understandable within one context window
  • Focused files: Prefer small, focused files

Output Format:

## File Structure

### New Files

- `path/to/new/file.ts` - Responsibility description

### Modified Files

- `path/to/existing/file.ts` - Modification content description

### Test Files

- `tests/path/to/test.ts` - Test responsibilities

Code Comment Planning:

Explicitly note comment requirements in file structure planning:

For each new or modified file, annotate comment requirements:

### New Files

- `src/utils/helper.ts` - Responsibility description
  - Comment requirements: Need documentation comments for all public functions
  - Complexity: Medium (contains complex algorithms, needs detailed comments)

Comment Requirements Evaluation Criteria:

  • High Priority: Public APIs, core business logic, complex algorithms
  • Medium Priority: Utility functions, data processing logic
  • Low Priority: Simple utility functions, configuration files

Completion Criteria: File structure planned, test files identified, comment requirements evaluated, user confirmed


Phase 3: Task Definition

Goal: Define specific implementation tasks.

Task Granularity: Each step should correspond to one operation (taking 2-5 minutes).

Task Structure:

### Task N: [Task Name]

**Files**:

- Create: `path/to/file.ts`
- Test: `tests/path/to/test.ts`

- [ ] **Step 1**: Write failing test
- [ ] **Step 2**: Run test to verify failure
- [ ] **Step 3**: Write minimal implementation
- [ ] **Step 4**: Run test to verify passing

Task Requirements:

  • Precise file paths, no vague references
  • Complete code in plan, no placeholders
  • Precise commands and expected output
  • Test-Driven Development (TDD): Each task must define test file and test cases first

Test Requirements:

⚠️ Tasks must include test steps, the following are prohibited:

  • ❌ Only write implementation code, no tests
  • ❌ Test file path is empty or uses placeholders
  • ❌ Test cases are incomplete or too simple
  • ❌ Skip "Write failing test" step

Test Coverage Standards:

  • Each function/method has at least one positive test case
  • Edge conditions have corresponding tests
  • Error handling paths have test coverage
  • Test files must be created before task starts

Completion Criteria: All tasks defined, each task includes complete test plan, user confirmed


Phase 4: Plan Save and Confirmation

Goal: Save plan document, obtain user approval.

Actions:

  1. Save plan to design/plans/YYYY-MM-DD-<feature>.md
  2. Request user to confirm plan content
  3. If user approves, ask if execution should begin

Standard Phrase:

"The plan is complete and saved. Please confirm the following:

Plan Document: design/plans/<filename>.md Total Tasks: [count]

Should I begin execution after confirmation?"


Phase 5: Execution Handover

Goal: Invoke executing-plans skill.

Invocation Format:

Invoke executing-plans skill, plan document path: <path>

Termination: After this phase completes, this skill ends.


Multi-Subsystem Handling

When multiple subsystems are identified:

1. Create Outline File

Path: design/plans/YYYY-MM-DD-<topic>-outline.md

Content:

# [Project Name] Implementation Plan Outline

## Subsystem Division

| Subsystem | Scope | Dependency | Status |
| ------- | ------ | ---- | ------ |
| SubsystemA | [scope] | None | To be written |
| SubsystemB | [scope] | A | To be written |

## Execution Order

1. SubsystemA (no dependencies)
2. SubsystemB (depends on A)

## Shared Components

| Component | Using Subsystems | Definition Location |
| -------- | ---------- | ------------- |
| [Component name] | A, B | In SubsystemA plan |

2. Write Plans for Each Subsystem

Each subsystem independently writes plan documents, update outline file status after completion.


Reference Guide

Reference DocumentPurpose
references/plan-template.mdPlan template
references/patterns.mdCommon plan patterns

FAQ

Plan needs modification?

Handling:

  1. Record changes in outline file
  2. Update affected plan documents
  3. Notify user of change scope

How to recover after dialogue interruption?

Handling:

  1. Read session state file design/.session-state.md
  2. Continue execution based on "Next Step"

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.