Spec Readiness Score: Check Your Hallucination Risk
Assess your project's spec coverage before AI coding. Score your PRD, API spec, schema, and architecture. Higher score = fewer hallucinations.
Spec Readiness Score: Check Your Hallucination Risk
TL;DR: Score your project's spec coverage before AI coding. Higher score = fewer hallucinations. Gaps in specs = specific failure modes.
How It Works
Rate your project on each spec category. Add up your points. Check your risk level.
The Scorecard
Category 1: Product Requirements (PRD)
| Criteria | Points | Have It? |
|---|---|---|
| Problem statement defined | 5 | ☐ |
| Target users identified | 5 | ☐ |
| Feature list (in scope) | 10 | ☐ |
| Non-goals (out of scope) | 10 | ☐ |
| Acceptance criteria per feature | 10 | ☐ |
| Success metrics defined | 5 | ☐ |
PRD Subtotal: ___ / 45
Category 2: API Specification
| Criteria | Points | Have It? |
|---|---|---|
| All endpoints listed | 10 | ☐ |
| Request bodies defined | 5 | ☐ |
| Response schemas defined | 5 | ☐ |
| Error responses defined | 5 | ☐ |
| Authentication documented | 5 | ☐ |
| Examples for each endpoint | 5 | ☐ |
API Subtotal: ___ / 35
Category 3: Database Schema
| Criteria | Points | Have It? |
|---|---|---|
| All tables documented | 10 | ☐ |
| All columns with types | 5 | ☐ |
| Relationships defined (FKs) | 5 | ☐ |
| Constraints documented | 3 | ☐ |
| Indexes documented | 2 | ☐ |
Schema Subtotal: ___ / 25
Category 4: Architecture
| Criteria | Points | Have It? |
|---|---|---|
| Component inventory | 5 | ☐ |
| Data flow diagram | 5 | ☐ |
| Integration points listed | 5 | ☐ |
| Tech stack documented | 5 | ☐ |
| Patterns/conventions defined | 5 | ☐ |
Architecture Subtotal: ___ / 25
Total Score
Add your subtotals:
| Category | Your Score | Max |
|---|---|---|
| PRD | ___ | 45 |
| API Spec | ___ | 35 |
| Schema | ___ | 25 |
| Architecture | ___ | 25 |
| TOTAL | ___ | 130 |
Risk Levels
🟢 100-130 Points: Low Risk
You're ready for AI coding.
Your specs are comprehensive. AI tools have the context they need.
Expected hallucination rate: < 10%
🟡 60-99 Points: Medium Risk
Proceed with caution.
Gaps exist. Expect some hallucinations in specific areas.
| Missing | Expect |
|---|---|
| PRD gaps | Scope creep, bonus features |
| API gaps | Wrong endpoints/payloads |
| Schema gaps | Wrong column names |
| Architecture gaps | Pattern conflicts |
Expected hallucination rate: 20-40%
🔴 0-59 Points: High Risk
Stop. Build specs first.
You're vibe coding. AI will invent constantly.
Expected hallucination rate: 50%+
Gap Analysis
Check which categories are lowest. Each gap causes specific failures:
| Low Score In | Hallucinations You'll See |
|---|---|
| PRD | AI adds features you didn't ask for |
| API Spec | AI invents endpoints, wrong payloads |
| Schema | AI queries non-existent columns |
| Architecture | AI mixes patterns, wrong integrations |
Improve Your Score
If PRD is Low
- Write a 1-page problem statement
- List 3-5 core features
- List 3-5 non-goals
- Add acceptance criteria per feature
If API Spec is Low
- List every endpoint you need
- Define request/response for each
- Add error responses
- Specify auth requirements
If Schema is Low
- List all tables
- Document all columns with types
- Mark required vs optional
- Define relationships
If Architecture is Low
- List all components (frontend, API, DB)
- Draw a simple data flow
- Document tech choices
- Define patterns to follow
→ Download Architecture Template
Generate Specs Automatically
Instead of writing all specs manually:
- Brain dump your project idea
- Answer 10-15 structured questions
- Generate 60+ documents automatically
Context Ark creates complete spec packs from a brain dump.
Related Resources
Templates
Guides
Last updated: January 2026
Context Ark Team
Writing about AI, documentation, and developer tools
Turn Brain Dumps into PRDs
Don't let AI guess your requirements. Generate a structured PRD with acceptance criteria instantly.