writing-plans · diff
git:20260501.3dcb621 to git:20261002.3a8121c
73 added, 118 removed. Audit A to A.
---
name: writing-plans
description: Use when you have a spec or requirements for a multi-step task, before touching code
---
- # Writing Plans
-
- ## Overview
-
- Write comprehensive implementation plans assuming the engineer has zero context for our codebase and questionable taste. Document everything they need to know: which files to touch for each task, code, testing, docs they might need to check, how to test it. Give them the whole plan as bite-sized tasks. DRY. YAGNI. TDD. Frequent commits.
-
- Assume they are a skilled developer, but know almost nothing about our toolset or problem domain. Assume they don't know good test design very well.
-
- **Announce at start:** "I'm using the writing-plans skill to create the implementation plan."
-
- **Context:** This should be run in a dedicated worktree (created by brainstorming skill).
-
- **Save plans to:** `docs/bearpaws/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.
-
- ## Bite-Sized Task Granularity
-
- **Each step is one action (2-5 minutes):**
- - "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 bp:subagent-driven-development (recommended) or bp: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]
-
- ---
- ```
-
- ## Task Structure
+ <skill>
- ````markdown
- ### Task N: [Component Name]
+ <purpose>
+ Write implementation plans assuming the engineer has zero codebase context and questionable taste. Document everything: which files to touch, code, testing, docs to check. Bite-sized tasks. DRY. YAGNI. TDD. Frequent commits. Assume a skilled developer who knows nothing about this toolset or domain.
+ </purpose>
- **Files:**
- - Create: `exact/path/to/file.py`
- - Modify: `exact/path/to/existing.py:123-145`
- - Test: `tests/exact/path/to/test.py`
+ <triggers>
+ <rule>Use after brainstorming produces an approved spec.</rule>
+ <rule>Use when you have requirements for a multi-step task, before writing code.</rule>
+ </triggers>
- - [ ] **Step 1: Write the failing test**
+ <warning level="soft">
+ Announce at start: "I'm using the writing-plans skill to create the implementation plan."
+ </warning>
- ```python
- def test_specific_behavior():
- result = function(input)
- assert result == expected
- ```
+ <rules>
+ <rule>**Save plans to:** `docs/bearpaws/plans/YYYY-MM-DD-{feature-name}.md` (user prefs override).</rule>
+ <rule>**In native planning mode** (the harness blocks project writes until the user approves a plan): build and self-review the plan in the harness's plan artifact. Do not leave native planning mode to save the plan. After approval, save it to the path above with `**Status:** Approved` before the first implementation change.</rule>
+ <rule>**Scope check:** If spec covers multiple independent subsystems, suggest separate plans — one per subsystem. Each plan produces working, testable software on its own.</rule>
+ </rules>
- - [ ] **Step 2: Run test to verify it fails**
+ <process>
+ <step>**Map file structure** — which files will be created or modified, what each is responsible for. Design units with clear boundaries. Prefer smaller focused files. Files that change together should live together.</step>
+ <step>**Define tasks** — each task produces self-contained changes. Each step is one action (2-5 min): write failing test, run it, implement minimal code, run tests, commit.</step>
+ <step>**Self-review** — (1) spec coverage: can you point to a task for every requirement? (2) placeholder scan: no TBD/TODO/vague steps. (3) type consistency: names in later tasks match definitions in earlier tasks. Fix inline.</step>
+ <step>**Execution handoff** — offer choice: **Subagent-Driven** (bp:subagent-driven-development, recommended) or **Inline** (bp:executing-plans).</step>
+ <step>**After native plan approval** — follow the native-approval gate below instead of offering a choice.</step>
+ </process>
- Run: `pytest tests/path/test.py::test_name -v`
- Expected: FAIL with "function not defined"
+ ## Plan document header
- - [ ] **Step 3: Write minimal implementation**
+ Every plan MUST start with:
- ```python
- def function(input):
- return expected
- ```
+ ```markdown
+ # [Feature Name] Implementation Plan
- - [ ] **Step 4: Run test to verify it passes**
+ > **For agentic workers:** REQUIRED SUB-SKILL: Use bp:subagent-driven-development (recommended) or bp:executing-plans to implement this plan task-by-task. Steps use checkbox syntax for tracking.
- Run: `pytest tests/path/test.py::test_name -v`
- Expected: PASS
+ **Goal:** [One sentence]
- - [ ] **Step 5: Commit**
+ **Architecture:** [2-3 sentences]
- ```bash
- git add tests/path/test.py src/path/file.py
- git commit -m "feat: add specific feature"
- ```
- ````
+ **Tech Stack:** [Key technologies/libraries]
- ## No Placeholders
+ ---
+ ```
- Every step must contain the actual content an engineer needs. These are **plan failures** — never write them:
- - "TBD", "TODO", "implement later", "fill in details"
- - "Add appropriate error handling" / "add validation" / "handle edge cases"
- - "Write tests for the above" (without actual test code)
- - "Similar to Task N" (repeat the code — the engineer may be reading tasks out of order)
- - Steps that describe what to do without showing how (code blocks required for code steps)
- - References to types, functions, or methods not defined in any task
+ ## Task structure
- ## Remember
- - Exact file paths always
- - Complete code in every step — if a step changes code, show the code
- - Exact commands with expected output
- - DRY, YAGNI, TDD, frequent commits
+ ````markdown
+ ### Task N: [Component Name]
- ## Self-Review
+ **Files:**
+ - Create: `exact/path/to/file.py`
+ - Modify: `exact/path/to/existing.py:123-145`
+ - Test: `tests/exact/path/to/test.py`
- 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.
+ - [ ] **Step 1: Write the failing test**
- **1. Spec coverage:** Skim each section/requirement in the spec. Can you point to a task that implements it? List any gaps.
+ ```python
+ def test_specific_behavior():
+ result = function(input)
+ assert result == expected
+ ```
- **2. Placeholder scan:** Search your plan for red flags — any of the patterns from the "No Placeholders" section above. Fix them.
+ - [ ] **Step 2: Run test — verify FAIL**
- **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.
+ Run: `pytest tests/path/test.py::test_name -v`
+ Expected: FAIL with "function not defined"
- 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.
+ - [ ] **Step 3: Write minimal implementation**
- ## Execution Handoff
+ ```python
+ def function(input):
+ return expected
+ ```
- After saving the plan, offer execution choice:
+ - [ ] **Step 4: Run test — verify PASS**
- **"Plan complete and saved to `docs/bearpaws/plans/<filename>.md`. Two execution options:**
+ Run: `pytest tests/path/test.py::test_name -v`
+ Expected: PASS
- **1. Subagent-Driven (recommended)** - I dispatch a fresh subagent per task, review between tasks, fast iteration
+ - [ ] **Step 5: Commit**
+ ````
- **2. Inline Execution** - Execute tasks in this session using executing-plans, batch execution with checkpoints
+ <gate name="native-approval">
+ Native plan approval is the go-ahead: do not ask again whether to start or how to execute. Before any code change, in order: (1) set up the workspace, (2) save the approved plan, (3) invoke `Skill` with bp:subagent-driven-development (tasks mostly independent) or bp:executing-plans. "I'll work through the plan inline" means invoking bp:executing-plans, not editing directly.
+ </gate>
- **Which approach?"**
+ <gate name="no-placeholders">
+ Every step must contain actual content. These are plan failures — never write them:
+ - "TBD", "TODO", "implement later", "fill in details"
+ - "Add appropriate error handling" / "add validation"
+ - "Write tests for the above" (without actual test code)
+ - "Similar to Task N" (repeat the code)
+ - Steps describing what to do without showing how
+ - References to types/functions not defined in any task
+ </gate>
- **If Subagent-Driven chosen:**
- - **REQUIRED SUB-SKILL:** Use bp:subagent-driven-development
- - Fresh subagent per task + two-stage review
+ <see file="plan-document-reviewer-prompt.md"/>
- **If Inline Execution chosen:**
- - **REQUIRED SUB-SKILL:** Use bp:executing-plans
- - Batch execution with checkpoints for review
+ </skill>