writing-plans

# Writing Plans

## Overview

Write implementation plans for an engineer who has not seen this codebase or this spec. Assume they write idiomatic code in the project's language once they know the exact interface and the exact test, and that they will make a reasonable choice wherever the plan leaves one open. What they cannot know is what you decided: which files, which names and signatures, which values from the spec, which tests prove each task. Document those. Give them the whole plan as bite-sized tasks. DRY. YAGNI. TDD. Frequent commits.

**Announce at start:** "I'm using the writing-plans skill to create the implementation plan."

**Context:** If working in an isolated worktree, it should have been created via the `superpowers:using-git-worktrees` skill at execution time.

**Save plans to:** `docs/superpowers/plans/YYYY-MM-DD-<feature-name>.md`
- (User preferences for plan location override this default)

## Scope Check

If the spec covers multiple independent subsystems, it should have been broken into sub-project specs during brainstorming. If it wasn't, suggest breaking this into separate plans — one per subsystem. Each plan should produce working, testable software on its own.

## File Structure

Before defining tasks, map out which files will be created or modified and what each one is responsible for. This is where decomposition decisions get locked in.

- Design units with clear boundaries and well-defined interfaces. Each file should have one clear responsibility.
- You reason best about code you can hold in context at once, and your edits are more reliable when files are focused. Prefer smaller, focused files over large ones that do too much.
- Files that change together should live together. Split by responsibility, not by technical layer.
- In existing codebases, follow established patterns. If the codebase uses large files, don't unilaterally restructure - but if a file you're modifying has grown unwieldy, including a split in the plan is reasonable.

This structure informs the task decomposition. Each task should produce self-contained changes that make sense independently.

## Task Right-Sizing

A task is the smallest unit that carries its own test cycle and is worth a
fresh reviewer's gate. When drawing task boundaries: fold setup,
configuration, scaffolding, and documentation steps into the task whose
deliverable needs them; split only where a reviewer could meaningfully
reject one task while approving its neighbor. Each task ends with an
independently testable deliverable.

## Step Granularity

**Each step is one action with a checkable result:**
- "Write the failing test" - step
- "Run it to make sure it fails" - step
- "Implement the minimal code to make the test pass" - step
- "Run the tests and make sure they pass" - step
- "Commit" - step

## Plan Document Header

**Every plan MUST start with this header:**

```markdown
# [Feature Name] Implementation Plan

> **For agentic workers:** REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use checkbox (`- [ ]`) syntax for tracking.

**Goal:** [One sentence describing what this builds]

**Architecture:** [2-3 sentences about approach]

**Tech Stack:** [Key technologies/libraries]

**Spec:** [path to the spec/design doc this plan implements — the plan
argues from the spec, so the spec travels with it; executors read both]

## Global Constraints

[The spec's project-wide requirements — version floors, dependency limits,
naming and copy rules, platform requirements — one line each, with exact
values copied verbatim from the spec. Every task's requirements implicitly
include this section.]

## Review Focus

[The five input classes or failure modes the spec implies but no task's
tests exercise that are most likely to bite a person using this software
— one line each, naming the input or condition and the behavior a
reasonable person would expect, most likely first. The spec is a vision
document: it says what the software must do, not everything it will
meet, and its silence on an input is not permission for that input to
break the program. Write the list here, once, with the spec in front of
you. Then, for each line, add the test that pins it to the task that
owns the code, in that task's own step style.]

---
```

## Task Structure

````markdown
### Task N: [Component Name]

**Files:**
- Create: `exact/path/to/file.py`
- Modify: `exact/path/to/existing.py:123-145`
- Test: `tests/exact/path/to/test.py`

**Interfaces:**
- Consumes: [what this task uses from earlier tasks — exact signatures]
- Produces: [what later tasks rely on — exact function names, parameter
  and return types. A task's implementer sees only their own task; this
  block is how they learn the names and types neighboring tasks use.]

- [ ] **Step 1: Write the failing test**

```python
def test_specific_behavior():
    result = function(input)
    assert result == expected
```

- [ ] **Step 2: Run test to verify it fails**

Run: `pytest tests/path/test.py::test_name -v`
Expected: FAIL with "function not defined"

- [ ] **Step 3: Implement `function(input: InputType) -> ResultType` in `exact/path/to/file.py`**

One line on the approach when the signature and the test leave a choice
(which library call, which data structure); a code block only for an
algorithm they do not determine.

- [ ] **Step 4: Run test to verify it passes**

Run: `pytest tests/path/test.py::test_name -v`
Expected: PASS

- [ ] **Step 5: Commit**

```bash
git add tests/path/test.py src/path/file.py
git commit -m "feat: add specific feature"
```
````

## What a Step Contains

A step is done when the implementer can write exactly one reasonable thing
from it. That is the whole requirement: unambiguous, not complete. Each kind
of step carries what makes it unambiguous and nothing more:

- **A test step:** the test's name and its assertions, as code, with the
  spec's exact values in them.
- **A code step:** the exact signature (name, parameters, return type), the
  file it lives in, and the specific values the spec pins. The implementer
  writes the body. A body appears only for an algorithm the signature and
  tests do not determine, or for exact copy the spec fixes.
- **A verification step:** the command to run and the output that means it
  passed.
- **A reference to another task:** that task's Interfaces block says what
  to use; the plan does not repeat that task's code.

A plan is the set of decisions the implementer cannot make alone. A plan
longer than the code it describes has written the code instead. Lines that
decide nothing ("TBD", "handle edge cases", "add appropriate validation",
"write tests for the above", a type or function no task defines) are the
opposite failure, and the self-review catches both.

## Self-Review

After writing the complete plan, look at the spec with fresh eyes and check the plan against it. This is a checklist you run yourself — not a subagent dispatch.

**1. Spec coverage:** Skim each section/requirement in the spec. Can you point to a task that implements it? List any gaps.

**2. Step scan:** Every step must let the implementer write exactly one reasonable thing, and no step may carry more than that: a line that decides nothing is a gap, a function body the signature and tests already determine is a transcript. Fix both.

**3. Type consistency:** Do the types, method signatures, and property names you used in later tasks match what you defined in earlier tasks? A function called `clearLayers()` in Task 3 but `clearFullLayers()` in Task 7 is a bug.

**4. Review Focus:** For each input class or failure mode the spec implies, is there a task whose tests exercise it? The five uncovered ones most likely to bite a person go in the Review Focus section, and each line there gets its test added to the owning task. An empty section means you checked and found none, not that you skipped the check.

**5. Proportion:** Compare the plan's length to the spec's. A plan several times longer than the spec it implements is a transcript of the program, not a plan. If code blocks are most of the document, replace bodies with signatures, test names and assertions, and check that each step is still unambiguous.

If you find issues, fix them inline. No need to re-review — just fix and move on. If you find a spec requirement with no task, add the task.

## Execution Handoff

After saving and self-reviewing the plan, link it for your human partner
to read. If they have already explicitly supplied an execution method, ask
them to review the plan and confirm it captures what they want; wait for that
review before implementation, then use the preserved method. Otherwise, ask
them to review the plan and choose an execution method before implementation.

**When no execution method has already been supplied:**

**"Plan complete and saved to `docs/superpowers/plans/<filename>.md`. Please review the plan. Which execution approach would you prefer?**

- **Subagent-driven** - A fresh subagent implements each task and a fresh reviewer checks it before the next one starts, then a whole-branch review at the end. Most thorough; costs a fresh context per task and per review.
- **Native** - I implement every task myself in this session, the way this harness runs work, then one fresh reviewer on the most capable model checks the whole branch. Cheapest and fastest; no independent review until the end. Runs well with a mid-tier session model, since the plan carries the design.

**For this plan I recommend <one of the two>, because <one sentence from the plan: how much the tasks depend on each other's interfaces, how many there are, what a shipped mistake would cost>. Does the plan capture what you want, and which approach should we use?"**

**When an execution method has already been supplied:**

**"Plan complete and saved to `docs/superpowers/plans/<filename>.md`. Please review the plan. Does it capture what you want?"**

**If Subagent-driven chosen:**
- **REQUIRED SUB-SKILL:** Use superpowers:subagent-driven-development

**If Native chosen:**
- **REQUIRED SUB-SKILL:** Use superpowers:executing-plans

{
  "format": "aiskill/v1",
  "skill": {
    "name": "writing-plans",
    "slug": "writing-plans",
    "version": "1.0.0",
    "description": "Use when you have a spec or requirements for a multi-step task, before touching code",
    "category": "Development",
    "tags": [
      "dev-workflow",
      "content"
    ],
    "platforms": [
      "chatgpt",
      "claude-code",
      "cursor"
    ],
    "tested_environment": "Agent Skills convention (SKILL.md) — works with Claude Code, Cursor, ChatGPT/Codex and compatible agents",
    "license": "MIT"
  },
  "prompts": {
    "system": "# Writing Plans\n\n## Overview\n\nWrite implementation plans for an engineer who has not seen this codebase or this spec. Assume they write idiomatic code in the project's language once they know the exact interface and the exact test, and that they will make a reasonable choice wherever the plan leaves one open. What they cannot know is what you decided: which files, which names and signatures, which values from the spec, which tests prove each task. Document those. Give them the whole plan as bite-sized tasks. DRY. YAGNI. TDD. Frequent commits.\n\n**Announce at start:** \"I'm using the writing-plans skill to create the implementation plan.\"\n\n**Context:** If working in an isolated worktree, it should have been created via the `superpowers:using-git-worktrees` skill at execution time.\n\n**Save plans to:** `docs/superpowers/plans/YYYY-MM-DD-<feature-name>.md`\n- (User preferences for plan location override this default)\n\n## Scope Check\n\nIf the spec covers multiple independent subsystems, it should have been broken into sub-project specs during brainstorming. If it wasn't, suggest breaking this into separate plans — one per subsystem. Each plan should produce working, testable software on its own.\n\n## File Structure\n\nBefore defining tasks, map out which files will be created or modified and what each one is responsible for. This is where decomposition decisions get locked in.\n\n- Design units with clear boundaries and well-defined interfaces. Each file should have one clear responsibility.\n- You reason best about code you can hold in context at once, and your edits are more reliable when files are focused. Prefer smaller, focused files over large ones that do too much.\n- Files that change together should live together. Split by responsibility, not by technical layer.\n- In existing codebases, follow established patterns. If the codebase uses large files, don't unilaterally restructure - but if a file you're modifying has grown unwieldy, including a split in the plan is reasonable.\n\nThis structure informs the task decomposition. Each task should produce self-contained changes that make sense independently.\n\n## Task Right-Sizing\n\nA task is the smallest unit that carries its own test cycle and is worth a\nfresh reviewer's gate. When drawing task boundaries: fold setup,\nconfiguration, scaffolding, and documentation steps into the task whose\ndeliverable needs them; split only where a reviewer could meaningfully\nreject one task while approving its neighbor. Each task ends with an\nindependently testable deliverable.\n\n## Step Granularity\n\n**Each step is one action with a checkable result:**\n- \"Write the failing test\" - step\n- \"Run it to make sure it fails\" - step\n- \"Implement the minimal code to make the test pass\" - step\n- \"Run the tests and make sure they pass\" - step\n- \"Commit\" - step\n\n## Plan Document Header\n\n**Every plan MUST start with this header:**\n\n```markdown\n# [Feature Name] Implementation Plan\n\n> **For agentic workers:** REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use checkbox (`- [ ]`) syntax for tracking.\n\n**Goal:** [One sentence describing what this builds]\n\n**Architecture:** [2-3 sentences about approach]\n\n**Tech Stack:** [Key technologies/libraries]\n\n**Spec:** [path to the spec/design doc this plan implements — the plan\nargues from the spec, so the spec travels with it; executors read both]\n\n## Global Constraints\n\n[The spec's project-wide requirements — version floors, dependency limits,\nnaming and copy rules, platform requirements — one line each, with exact\nvalues copied verbatim from the spec. Every task's requirements implicitly\ninclude this section.]\n\n## Review Focus\n\n[The five input classes or failure modes the spec implies but no task's\ntests exercise that are most likely to bite a person using this software\n— one line each, naming the input or condition and the behavior a\nreasonable person would expect, most likely first. The spec is a vision\ndocument: it says what the software must do, not everything it will\nmeet, and its silence on an input is not permission for that input to\nbreak the program. Write the list here, once, with the spec in front of\nyou. Then, for each line, add the test that pins it to the task that\nowns the code, in that task's own step style.]\n\n---\n```\n\n## Task Structure\n\n````markdown\n### Task N: [Component Name]\n\n**Files:**\n- Create: `exact/path/to/file.py`\n- Modify: `exact/path/to/existing.py:123-145`\n- Test: `tests/exact/path/to/test.py`\n\n**Interfaces:**\n- Consumes: [what this task uses from earlier tasks — exact signatures]\n- Produces: [what later tasks rely on — exact function names, parameter\n  and return types. A task's implementer sees only their own task; this\n  block is how they learn the names and types neighboring tasks use.]\n\n- [ ] **Step 1: Write the failing test**\n\n```python\ndef test_specific_behavior():\n    result = function(input)\n    assert result == expected\n```\n\n- [ ] **Step 2: Run test to verify it fails**\n\nRun: `pytest tests/path/test.py::test_name -v`\nExpected: FAIL with \"function not defined\"\n\n- [ ] **Step 3: Implement `function(input: InputType) -> ResultType` in `exact/path/to/file.py`**\n\nOne line on the approach when the signature and the test leave a choice\n(which library call, which data structure); a code block only for an\nalgorithm they do not determine.\n\n- [ ] **Step 4: Run test to verify it passes**\n\nRun: `pytest tests/path/test.py::test_name -v`\nExpected: PASS\n\n- [ ] **Step 5: Commit**\n\n```bash\ngit add tests/path/test.py src/path/file.py\ngit commit -m \"feat: add specific feature\"\n```\n````\n\n## What a Step Contains\n\nA step is done when the implementer can write exactly one reasonable thing\nfrom it. That is the whole requirement: unambiguous, not complete. Each kind\nof step carries what makes it unambiguous and nothing more:\n\n- **A test step:** the test's name and its assertions, as code, with the\n  spec's exact values in them.\n- **A code step:** the exact signature (name, parameters, return type), the\n  file it lives in, and the specific values the spec pins. The implementer\n  writes the body. A body appears only for an algorithm the signature and\n  tests do not determine, or for exact copy the spec fixes.\n- **A verification step:** the command to run and the output that means it\n  passed.\n- **A reference to another task:** that task's Interfaces block says what\n  to use; the plan does not repeat that task's code.\n\nA plan is the set of decisions the implementer cannot make alone. A plan\nlonger than the code it describes has written the code instead. Lines that\ndecide nothing (\"TBD\", \"handle edge cases\", \"add appropriate validation\",\n\"write tests for the above\", a type or function no task defines) are the\nopposite failure, and the self-review catches both.\n\n## Self-Review\n\nAfter writing the complete plan, look at the spec with fresh eyes and check the plan against it. This is a checklist you run yourself — not a subagent dispatch.\n\n**1. Spec coverage:** Skim each section/requirement in the spec. Can you point to a task that implements it? List any gaps.\n\n**2. Step scan:** Every step must let the implementer write exactly one reasonable thing, and no step may carry more than that: a line that decides nothing is a gap, a function body the signature and tests already determine is a transcript. Fix both.\n\n**3. Type consistency:** Do the types, method signatures, and property names you used in later tasks match what you defined in earlier tasks? A function called `clearLayers()` in Task 3 but `clearFullLayers()` in Task 7 is a bug.\n\n**4. Review Focus:** For each input class or failure mode the spec implies, is there a task whose tests exercise it? The five uncovered ones most likely to bite a person go in the Review Focus section, and each line there gets its test added to the owning task. An empty section means you checked and found none, not that you skipped the check.\n\n**5. Proportion:** Compare the plan's length to the spec's. A plan several times longer than the spec it implements is a transcript of the program, not a plan. If code blocks are most of the document, replace bodies with signatures, test names and assertions, and check that each step is still unambiguous.\n\nIf you find issues, fix them inline. No need to re-review — just fix and move on. If you find a spec requirement with no task, add the task.\n\n## Execution Handoff\n\nAfter saving and self-reviewing the plan, link it for your human partner\nto read. If they have already explicitly supplied an execution method, ask\nthem to review the plan and confirm it captures what they want; wait for that\nreview before implementation, then use the preserved method. Otherwise, ask\nthem to review the plan and choose an execution method before implementation.\n\n**When no execution method has already been supplied:**\n\n**\"Plan complete and saved to `docs/superpowers/plans/<filename>.md`. Please review the plan. Which execution approach would you prefer?**\n\n- **Subagent-driven** - A fresh subagent implements each task and a fresh reviewer checks it before the next one starts, then a whole-branch review at the end. Most thorough; costs a fresh context per task and per review.\n- **Native** - I implement every task myself in this session, the way this harness runs work, then one fresh reviewer on the most capable model checks the whole branch. Cheapest and fastest; no independent review until the end. Runs well with a mid-tier session model, since the plan carries the design.\n\n**For this plan I recommend <one of the two>, because <one sentence from the plan: how much the tasks depend on each other's interfaces, how many there are, what a shipped mistake would cost>. Does the plan capture what you want, and which approach should we use?\"**\n\n**When an execution method has already been supplied:**\n\n**\"Plan complete and saved to `docs/superpowers/plans/<filename>.md`. Please review the plan. Does it capture what you want?\"**\n\n**If Subagent-driven chosen:**\n- **REQUIRED SUB-SKILL:** Use superpowers:subagent-driven-development\n\n**If Native chosen:**\n- **REQUIRED SUB-SKILL:** Use superpowers:executing-plans",
    "user_template": ""
  },
  "variables": [],
  "install": {
    "steps": [
      {
        "title": "Download & unzip",
        "detail": "Download the .zip package from SkillVault and unzip it. The writing-plans/ folder contains SKILL.md plus any companion scripts, references and assets."
      },
      {
        "title": "Copy into your agent skills directory",
        "detail": "Move the folder into your agent's skills directory — e.g. ~/.claude/skills/ (Claude Code), .codex/skills/, .cursor/skills/ or your project's .claude/skills/ folder — then restart or reload the session."
      },
      {
        "title": "Use it",
        "detail": "Describe the task in natural language; the agent loads the skill automatically when the request matches the skill description."
      }
    ]
  },
  "usage": {
    "steps": []
  },
  "metadata": {
    "author": "SkillVault",
    "source": "https://skillvault.top/skills/writing-plans",
    "exported_at": "2026-11-03T07:43:12.143Z"
  }
}

format: aiskill/v1
skill:
  name: writing-plans
  slug: writing-plans
  version: 1.0.0
  description: Use when you have a spec or requirements for a multi-step task, before touching code
  category: Development
  tags:
    - dev-workflow
    - content
  platforms:
    - chatgpt
    - claude-code
    - cursor
  tested_environment: >-
    Agent Skills convention (SKILL.md) — works with Claude Code, Cursor, ChatGPT/Codex and
    compatible agents
  license: MIT
prompts:
  system: >-
    # Writing Plans


    ## Overview


    Write implementation plans for an engineer who has not seen this codebase or this spec. Assume
    they write idiomatic code in the project's language once they know the exact interface and the
    exact test, and that they will make a reasonable choice wherever the plan leaves one open. What
    they cannot know is what you decided: which files, which names and signatures, which values from
    the spec, which tests prove each task. Document those. Give them the whole plan as bite-sized
    tasks. DRY. YAGNI. TDD. Frequent commits.


    **Announce at start:** "I'm using the writing-plans skill to create the implementation plan."


    **Context:** If working in an isolated worktree, it should have been created via the
    `superpowers:using-git-worktrees` skill at execution time.


    **Save plans to:** `docs/superpowers/plans/YYYY-MM-DD-<feature-name>.md`

    - (User preferences for plan location override this default)


    ## Scope Check


    If the spec covers multiple independent subsystems, it should have been broken into sub-project
    specs during brainstorming. If it wasn't, suggest breaking this into separate plans — one per
    subsystem. Each plan should produce working, testable software on its own.


    ## File Structure


    Before defining tasks, map out which files will be created or modified and what each one is
    responsible for. This is where decomposition decisions get locked in.


    - Design units with clear boundaries and well-defined interfaces. Each file should have one
    clear responsibility.

    - You reason best about code you can hold in context at once, and your edits are more reliable
    when files are focused. Prefer smaller, focused files over large ones that do too much.

    - Files that change together should live together. Split by responsibility, not by technical
    layer.

    - In existing codebases, follow established patterns. If the codebase uses large files, don't
    unilaterally restructure - but if a file you're modifying has grown unwieldy, including a split
    in the plan is reasonable.


    This structure informs the task decomposition. Each task should produce self-contained changes
    that make sense independently.


    ## Task Right-Sizing


    A task is the smallest unit that carries its own test cycle and is worth a

    fresh reviewer's gate. When drawing task boundaries: fold setup,

    configuration, scaffolding, and documentation steps into the task whose

    deliverable needs them; split only where a reviewer could meaningfully

    reject one task while approving its neighbor. Each task ends with an

    independently testable deliverable.


    ## Step Granularity


    **Each step is one action with a checkable result:**

    - "Write the failing test" - step

    - "Run it to make sure it fails" - step

    - "Implement the minimal code to make the test pass" - step

    - "Run the tests and make sure they pass" - step

    - "Commit" - step


    ## Plan Document Header


    **Every plan MUST start with this header:**


    ```markdown

    # [Feature Name] Implementation Plan


    > **For agentic workers:** REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development
    (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use
    checkbox (`- [ ]`) syntax for tracking.


    **Goal:** [One sentence describing what this builds]


    **Architecture:** [2-3 sentences about approach]


    **Tech Stack:** [Key technologies/libraries]


    **Spec:** [path to the spec/design doc this plan implements — the plan

    argues from the spec, so the spec travels with it; executors read both]


    ## Global Constraints


    [The spec's project-wide requirements — version floors, dependency limits,

    naming and copy rules, platform requirements — one line each, with exact

    values copied verbatim from the spec. Every task's requirements implicitly

    include this section.]


    ## Review Focus


    [The five input classes or failure modes the spec implies but no task's

    tests exercise that are most likely to bite a person using this software

    — one line each, naming the input or condition and the behavior a

    reasonable person would expect, most likely first. The spec is a vision

    document: it says what the software must do, not everything it will

    meet, and its silence on an input is not permission for that input to

    break the program. Write the list here, once, with the spec in front of

    you. Then, for each line, add the test that pins it to the task that

    owns the code, in that task's own step style.]


    ---

    ```


    ## Task Structure


    ````markdown

    ### Task N: [Component Name]


    **Files:**

    - Create: `exact/path/to/file.py`

    - Modify: `exact/path/to/existing.py:123-145`

    - Test: `tests/exact/path/to/test.py`


    **Interfaces:**

    - Consumes: [what this task uses from earlier tasks — exact signatures]

    - Produces: [what later tasks rely on — exact function names, parameter
      and return types. A task's implementer sees only their own task; this
      block is how they learn the names and types neighboring tasks use.]

    - [ ] **Step 1: Write the failing test**


    ```python

    def test_specific_behavior():
        result = function(input)
        assert result == expected
    ```


    - [ ] **Step 2: Run test to verify it fails**


    Run: `pytest tests/path/test.py::test_name -v`

    Expected: FAIL with "function not defined"


    - [ ] **Step 3: Implement `function(input: InputType) -> ResultType` in
    `exact/path/to/file.py`**


    One line on the approach when the signature and the test leave a choice

    (which library call, which data structure); a code block only for an

    algorithm they do not determine.


    - [ ] **Step 4: Run test to verify it passes**


    Run: `pytest tests/path/test.py::test_name -v`

    Expected: PASS


    - [ ] **Step 5: Commit**


    ```bash

    git add tests/path/test.py src/path/file.py

    git commit -m "feat: add specific feature"

    ```

    ````


    ## What a Step Contains


    A step is done when the implementer can write exactly one reasonable thing

    from it. That is the whole requirement: unambiguous, not complete. Each kind

    of step carries what makes it unambiguous and nothing more:


    - **A test step:** the test's name and its assertions, as code, with the
      spec's exact values in them.
    - **A code step:** the exact signature (name, parameters, return type), the
      file it lives in, and the specific values the spec pins. The implementer
      writes the body. A body appears only for an algorithm the signature and
      tests do not determine, or for exact copy the spec fixes.
    - **A verification step:** the command to run and the output that means it
      passed.
    - **A reference to another task:** that task's Interfaces block says what
      to use; the plan does not repeat that task's code.

    A plan is the set of decisions the implementer cannot make alone. A plan

    longer than the code it describes has written the code instead. Lines that

    decide nothing ("TBD", "handle edge cases", "add appropriate validation",

    "write tests for the above", a type or function no task defines) are the

    opposite failure, and the self-review catches both.


    ## Self-Review


    After writing the complete plan, look at the spec with fresh eyes and check the plan against it.
    This is a checklist you run yourself — not a subagent dispatch.


    **1. Spec coverage:** Skim each section/requirement in the spec. Can you point to a task that
    implements it? List any gaps.


    **2. Step scan:** Every step must let the implementer write exactly one reasonable thing, and no
    step may carry more than that: a line that decides nothing is a gap, a function body the
    signature and tests already determine is a transcript. Fix both.


    **3. Type consistency:** Do the types, method signatures, and property names you used in later
    tasks match what you defined in earlier tasks? A function called `clearLayers()` in Task 3 but
    `clearFullLayers()` in Task 7 is a bug.


    **4. Review Focus:** For each input class or failure mode the spec implies, is there a task
    whose tests exercise it? The five uncovered ones most likely to bite a person go in the Review
    Focus section, and each line there gets its test added to the owning task. An empty section
    means you checked and found none, not that you skipped the check.


    **5. Proportion:** Compare the plan's length to the spec's. A plan several times longer than the
    spec it implements is a transcript of the program, not a plan. If code blocks are most of the
    document, replace bodies with signatures, test names and assertions, and check that each step is
    still unambiguous.


    If you find issues, fix them inline. No need to re-review — just fix and move on. If you find a
    spec requirement with no task, add the task.


    ## Execution Handoff


    After saving and self-reviewing the plan, link it for your human partner

    to read. If they have already explicitly supplied an execution method, ask

    them to review the plan and confirm it captures what they want; wait for that

    review before implementation, then use the preserved method. Otherwise, ask

    them to review the plan and choose an execution method before implementation.


    **When no execution method has already been supplied:**


    **"Plan complete and saved to `docs/superpowers/plans/<filename>.md`. Please review the plan.
    Which execution approach would you prefer?**


    - **Subagent-driven** - A fresh subagent implements each task and a fresh reviewer checks it
    before the next one starts, then a whole-branch review at the end. Most thorough; costs a fresh
    context per task and per review.

    - **Native** - I implement every task myself in this session, the way this harness runs work,
    then one fresh reviewer on the most capable model checks the whole branch. Cheapest and fastest;
    no independent review until the end. Runs well with a mid-tier session model, since the plan
    carries the design.


    **For this plan I recommend <one of the two>, because <one sentence from the plan: how much the
    tasks depend on each other's interfaces, how many there are, what a shipped mistake would cost>.
    Does the plan capture what you want, and which approach should we use?"**


    **When an execution method has already been supplied:**


    **"Plan complete and saved to `docs/superpowers/plans/<filename>.md`. Please review the plan.
    Does it capture what you want?"**


    **If Subagent-driven chosen:**

    - **REQUIRED SUB-SKILL:** Use superpowers:subagent-driven-development


    **If Native chosen:**

    - **REQUIRED SUB-SKILL:** Use superpowers:executing-plans
  user_template: ""
variables: []
install:
  steps:
    - title: Download & unzip
      detail: >-
        Download the .zip package from SkillVault and unzip it. The writing-plans/ folder contains
        SKILL.md plus any companion scripts, references and assets.
    - title: Copy into your agent skills directory
      detail: >-
        Move the folder into your agent's skills directory — e.g. ~/.claude/skills/ (Claude Code),
        .codex/skills/, .cursor/skills/ or your project's .claude/skills/ folder — then restart or
        reload the session.
    - title: Use it
      detail: >-
        Describe the task in natural language; the agent loads the skill automatically when the
        request matches the skill description.
usage:
  steps: []
metadata:
  author: SkillVault
  source: https://skillvault.top/skills/writing-plans
  exported_at: "2026-11-03T07:43:12.143Z"