Feature Spec Author

VerifiedSafe

Transforms a PRD into a detailed engineering specification for developers.

Sby Skills Guide Bot
DocumentationIntermediate
7/25/2026
Claude Code
#engineering-spec#prd#technical-writing#architecture#planning

Recommended for

Automated summary

Turns an approved PRD into a structured eight-section engineering spec that is ready to guide implementation.

Strengths

  • Enforces a complete 8-section template: architecture, data models, API design, implementation plan, testing, risks and open questions
  • The implementation plan is an ordered task table with dependencies and size estimates
  • Always includes a risk/mitigation table and an open-questions list to surface PRD ambiguities
  • Outputs Markdown that can be committed and reviewed alongside the codebase

Limitations

  • Output quality depends entirely on how precise the input PRD is
  • Only does shallow reads of existing code through Read/Glob, so real system impact is not verified
  • No automated validation of the generated document (no linting or consistency check against the code)
When to use it

Use after a PRD is approved, to draft the technical specification for a feature before implementation starts.

When not to use it

Avoid it when the product requirements are still unstable, or for small fixes that don't warrant a full design document.

Security analysis

Safe
Quality score86/100

The skill only uses Read, Glob, and Write to read a PRD file and produce a Markdown engineering spec at a caller-specified path. There is no bash execution, no network access, no credential or secret handling, no obfuscated payloads, and no instructions to disable safety controls. The only mild consideration is that it writes to a path supplied in context, which is inherent to its documented purpose and carries no escalation risk beyond normal file authoring.

No concerns found

How this analysis is made

Examples

Spec from an approved PRD
Read docs/prd/user-authentication.md and write the full engineering spec to docs/specs/user-authentication-spec.md following the 8-section format.
Update an existing spec
The PRD at docs/prd/billing.md was updated with a new refund flow. Read it and update docs/specs/billing-spec.md, adding the new data models, API endpoints and implementation tasks.
Surface open questions
Draft the engineering spec for the notification-preferences PRD at docs/prd/notifications.md and list every ambiguity as an open question at the end.

name: feature-spec-author description: > Draft detailed engineering specs from an approved PRD. Use when generating or updating the engineering specification for a feature, before implementation begins. allowed-tools: Read,Glob,Write

Feature Spec Author

You are an expert software architect tasked with transforming a Product Requirements Document (PRD) into a detailed engineering specification.

Instructions

  1. Read the PRD at the path provided in the context
  2. Analyze the requirements, constraints, and success criteria thoroughly
  3. Generate a comprehensive engineering spec covering all sections below
  4. Write the output to the specified output path

Analysis Process

Before writing the spec, analyze:

  1. Core Requirements - What must be built?
  2. User Flows - How will users interact with the feature?
  3. Data Requirements - What data needs to be stored, transformed, or displayed?
  4. Integration Points - What existing systems does this touch?
  5. Edge Cases - What could go wrong?
  6. Non-Functional Requirements - Performance, security, scalability

Output Format

Write a Markdown document with these sections:

# Engineering Spec: [Feature Name]

## 1. Overview

### 1.1 Purpose
Brief description of what this feature does and why.

### 1.2 Scope
What is in scope and out of scope for this implementation.

### 1.3 Success Criteria
Measurable criteria for feature completion.

## 2. Architecture

### 2.1 High-Level Design
System architecture overview, diagrams if helpful.

### 2.2 Components
List of new or modified components with responsibilities.

### 2.3 Data Flow
How data moves through the system.

## 3. Data Models

### 3.1 New Models
```python
@dataclass
class NewModel:
    field: type  # Description

3.2 Schema Changes

Database migrations or schema updates needed.

4. API Design

4.1 Endpoints

| Method | Path | Description | |--------|------|-------------| | POST | /api/v1/resource | Create resource |

4.2 Request/Response Schemas

{
  "field": "type"
}

5. Implementation Plan

5.1 Tasks

Ordered list of implementation tasks with dependencies.

| # | Task | Dependencies | Size | |---|------|--------------|------| | 1 | Create data models | None | S |

5.2 File Changes

List of files to create or modify.

6. Testing Strategy

6.1 Unit Tests

Key unit tests needed.

6.2 Integration Tests

Integration test scenarios.

6.3 Edge Cases

Edge cases to cover.

7. Risks and Mitigations

| Risk | Impact | Mitigation | |------|--------|------------| | Risk description | High/Med/Low | How to mitigate |

8. Open Questions

Questions that need answers before or during implementation.


## Quality Standards

Your spec should:
- Be **actionable** - developers can implement directly from it
- Be **complete** - cover all aspects of the PRD
- Be **clear** - no ambiguous requirements
- Include **concrete examples** where helpful
- Identify **all dependencies** between tasks
- Surface **risks** proactively

## Example

Given a PRD for "User Authentication":

```markdown
# Engineering Spec: User Authentication

## 1. Overview

### 1.1 Purpose
Implement secure user authentication allowing users to sign up, log in, and manage their sessions.

### 1.2 Scope
**In Scope:**
- Email/password authentication
- Session management with JWT tokens
- Password reset flow

**Out of Scope:**
- Social authentication (OAuth)
- Multi-factor authentication

### 1.3 Success Criteria
- Users can create accounts and log in
- Sessions expire after 24 hours of inactivity
- Failed login attempts are rate-limited
...

Important Notes

  • Focus on the how, not the what (the PRD covers what)
  • Be specific about data types and validation rules
  • Include error handling requirements
  • Consider backwards compatibility if modifying existing systems
  • Flag any ambiguities in the PRD as open questions

From the file to a first result

The download contains exactly the content displayed on this page. This copy may differ from the latest version in the source repository.

View the source and supporting files

This download contains SKILL.md only. Any scripts, references, templates and dependencies are not included. Get the required files from the source, preserving their directory structure, before continuing.

1. Download the file

Download SKILL.md

From the root of a test project, save the file as .skill-downloads/feature-spec-author-mrzxuqp7/SKILL.md, or use the command below. This folder is for inspecting the skill before installation.

Command for macOS/Linux or WSL: requires a POSIX shell and curl. On Windows without WSL, use the download button and create the folder manually.

mkdir -p '.skill-downloads/feature-spec-author-mrzxuqp7' && curl --fail --show-error --location 'https://skills-guides.com/api/skills/feature-spec-author-mrzxuqp7/download' --output '.skill-downloads/feature-spec-author-mrzxuqp7/SKILL.md.part' && mv '.skill-downloads/feature-spec-author-mrzxuqp7/SKILL.md.part' '.skill-downloads/feature-spec-author-mrzxuqp7/SKILL.md'

2. Check prerequisites

Read SKILL.md before using it. If it references scripts or external services, inspect them and prepare the required access in your test environment.

Tools declared by the author (check availability in your agent): Read,Glob,Write

No recognized prerequisites section in this file. This does not mean it is self-contained: libraries, tools and access requirements may be mentioned elsewhere in the content or source repository.

The platforms on this page are catalog indications. The paths below follow tool documentation; they do not prove this skill has been run in those tools. No minimum tool version is confirmed for this listing.

3. Load the skill

After inspection, copy the complete folder into your project using one of the commands below, then open a new tool session. The command stops if a folder with the same name already exists: compare the files before replacing it.

Claude Code
test -f '.skill-downloads/feature-spec-author-mrzxuqp7/SKILL.md' && test -s '.skill-downloads/feature-spec-author-mrzxuqp7/SKILL.md' && mkdir -p '.claude/skills' && test ! -e '.claude/skills/feature-spec-author' && cp -R '.skill-downloads/feature-spec-author-mrzxuqp7' '.claude/skills/feature-spec-author'

OpenSkills

OpenSkills must already be installed. The local install/read procedure is checked with OpenSkills 1.5.0, which requires Node.js ≥ 20.6.0. Catalog slugs are not resolved automatically.

Official OpenSkills installation
openskills install './.skill-downloads/feature-spec-author-mrzxuqp7'

openskills read 'feature-spec-author-mrzxuqp7'

Reading should display the installed file content and its base directory. It does not run the skill or install its dependencies.

For an agent that reads AGENTS.md, then sync the index. This command changes AGENTS.md: review the diff before using it.

openskills sync

Manual method or another tool: attach SKILL.md to a conversation, or open it in an agent that can read files. Include any required supporting resources. No universal folder is recognized by every tool.

4. Try a first task

Use a short, synthetic input: an API endpoint, a small document or an interface brief. In a new conversation, attach the file or give the agent its path, then adapt this request:

Read .skill-downloads/feature-spec-author-mrzxuqp7/SKILL.md and apply its instructions to this example: [describe a test input]. Identify any missing files, dependencies or access. Produce the result requested by the skill without inventing data or running unreviewed scripts.

Expected result: output matching the skill instructions, or a precise list of missing prerequisites. A read confirmation alone does not validate a first use. If an external resource is unavailable, that path remains untested.

Related skills