writing-plans

Create implementation plans for an engineer with zero codebase context.

Each plan includes:

• Exact file paths for every operation
• Complete code (not "add validation here")
• Test-first approach with verification commands
• Bite-sized steps (2-5 min each)

Principles: DRY, YAGNI, TDD, frequent commits.

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

Context: Run in dedicated worktree. If none exists, use using-git-worktrees skill first.

Save plans to: docs/plans/YYYY-MM-DD-.md

1. Read spec/requirements completely
2. Explore project structure (view .)
3. Identify tech stack (package.json, pyproject.toml, etc.)
4. Note existing patterns in similar files
5. Check docs/ for existing conventions

Each step is one action (2-5 minutes), independently verifiable:

• "Write the failing test" — step
• "Run it to confirm failure" — step
• "Implement minimal code to pass" — step
• "Run tests to confirm pass" — step
• "Commit" — step

Every plan MUST start with this header:

~~~markdown

# [Feature Name] Implementation Plan

Goal: [One sentence describing what this builds]

Architecture: [2-3 sentences about approach]

Tech Stack: [Key technologies/libraries]

---

~~~

~~~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

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: Write minimal implementation

```python

def function(input):

return expected

```

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"

```

~~~

Verify plan completeness:

• Every file path exists or will be created
• Every command can be run exactly as written
• No TODO/placeholder text remains
• Tests cover all acceptance criteria from spec
• Include exact test code, not descriptions

After saving plan, present:

"Plan saved to docs/plans/.md. Choose execution mode:

1. Subagent-Driven — same session, fresh subagent per task, fast iteration
2. Parallel Session — new session, batched execution with checkpoints

Which approach?"

If Subagent-Driven chosen

• Stay in this session
• REQUIRED SUB-SKILL: subagent-driven-development
• Fresh subagent per task + two-stage review

If Parallel Session chosen

• Guide user to open new session in worktree
• REQUIRED SUB-SKILL: New session uses executing-plans