From tiny-brain
Add a feature to an existing PRD. Use when user wants to add functionality to an existing product plan.
How this skill is triggered — by the user, by Claude, or both
Slash command
/tiny-brain:featureThis skill is limited to the following tools:
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
Add a feature when the user wants to:
Add a feature when the user wants to:
Ask the user which PRD to add the feature to, or detect from context.
List existing PRDs:
ls docs/prd/
Work with the user to define:
Use the template at templates/feature-template.md.
Save to: docs/prd/{prd-id}/features/{feature-id}.md
YAML Frontmatter:
---
id: feature-kebab-case-id
prd_id: parent-prd-id
number: N # Next number after existing features (check docs/prd/{prd-id}/features/)
title: Feature Title
status: defined
created: YYYY-MM-DD
updated: YYYY-MM-DD
---
Feature Numbering: The number field determines task ID prefixes (task-{number}-{n}) and feature ordering. Check existing features in the PRD to find the next available number.
Critical: Use this exact format for task extraction:
## Tasks
### 1. First task title
Description of what needs to be done.
**Files to modify/create:**
- `path/to/file1.ts`
- `path/to/file2.ts`
**Expected changes:**
- Change 1
- Change 2
### 2. Second task title
Description...
**Files to modify/create:**
- `path/to/file.ts`
Each ### N. Title becomes a trackable task with:
test: prefix)feat: prefix)refactor: prefix)Task Granularity Guidance:
Task: header in the commit message for trackingTask: headers in one commit will all be tracked with the same commit SHAExample multi-task commit:
feat(dashboard): add hooks display components
PRD: dashboard-hooks-display
Feature: hooks-display-ui
Task: Add hooks state to RepoDetailPage
Task: Add Hooks tab to repo header
Task: Create HooksList component
Implements hooks list and selection...
IMPORTANT: Before implementation, analyze the codebase to identify relevant tests.
Find existing tests for affected code:
__tests__/ directoriesRead the test files to understand:
Categorize each test:
Add a Test Plan section using emoji categories:
| Emoji | Category | Description |
|---|---|---|
🔒 | Regression | Must pass unchanged |
✏️ | Amended Case | Existing case to be modified |
🆕 | New Case | New test case in existing file |
📄 | New File | Entirely new test file |
## Test Plan
### 🔒 Regression Tests
| File | Cases | Status |
|------|-------|--------|
| src/__tests__/existing.test.ts | all existing | ❌ |
### 🆕 New Tests
| File | Case | Status |
|------|------|--------|
| src/__tests__/feature.test.ts | handles new feature | ❌ |
Add a reference to the new feature in docs/prd/{prd-id}/prd.md:
### Feature N: {Feature Title}
**File**: [features/{feature-id}.md](features/{feature-id}.md)
**Status**: defined
**Description**: Brief description of what this feature does
After creating the feature file, call:
plan sync
This updates progress.json with the new feature and tasks.
Tell the user:
"I've added feature '{title}' to PRD '{prd-id}' with {N} tasks."
### N. Task Name format (numbered)templates/feature-template.mdUser: "Add a quality CLI command to the code-quality PRD"
Claude:
1. Confirm PRD: "Adding to code-quality-analysis PRD?"
2. Discuss: "What subcommands? What options?"
3. Create:
- docs/prd/code-quality-analysis/features/quality-cli-command.md
4. Update:
- docs/prd/code-quality-analysis/prd.md (add feature reference)
5. Call `plan sync`
6. Confirm: "Added 'Quality CLI Command' with 5 tasks"
npx claudepluginhub magic-ingredients/tiny-brain-releases --plugin tiny-brainExtends existing PRDs with new requirements via interactive clarification, generates sequential tasks, processes file paths and URLs as context. For iterative feature planning.
Generates feature-specific FRDs via interactive session prompting for context, user stories, functional requirements, technical specs, acceptance criteria, and saves as Markdown file.
Generates Product Requirements Documents (PRDs) for new features via phased approach: discovery, codebase exploration with agents, clarifying questions. Useful for feature planning or project starts.