Resources
Reference & Documentation
Guides and case studies are being prepared. In the meantime, this page documents the review engine itself — the five finding categories, scoring calibration, and how each piece of the review works.
Guides & case studies are being prepared
How-to guides, worked examples, and real-world case studies are in progress. Sign up to be notified when they publish.
Finding Categories
The Five Review Categories
Every finding in a Requirements Forge review belongs to one of five categories. Only Required Clarifications reduce the score — everything else is informational or advisory.
Required Clarification
Directly blocks implementation or testing. Reduces the review score.
Critical
Developer cannot proceed without guessing. Examples: no acceptance criteria at all, undefined core business rule, conflicting rules.
High
Significant gap that will cause rework. Examples: ambiguous condition, missing validation rule for a key field, undefined term used throughout.
Medium
Creates uncertainty but may be resolvable with assumptions. Examples: missing edge case, unclear secondary error behavior.
Question
Open questions a BA would put to the stakeholder. Split by urgency.
Required
Must be answered before development can begin.
Recommended
Helpful context. Development could proceed with a reasonable assumption.
Strength
Well-written aspects worth highlighting. No score impact. Examples: clear acceptance criteria, well-defined actor/trigger/outcome, good edge case coverage.
Recommendation
Optional improvements that do not materially reduce the score. Examples: business value statement, performance target, accessibility consideration, audit suggestion.
Solution Design
Technical concerns for architects and developers — never the BA's job and never reduces the score. Examples: optimistic locking, API design, database structure, caching strategy.
Scoring
How the Readiness Score Works
The score reflects one thing: implementation readiness. Only Required Clarifications reduce it. Recommendations, Solution Design findings, and Questions contribute zero to score reduction.
85–95
Ready for Development
Clear, well-written functional requirement with defined actor, trigger, outcome, and testable acceptance criteria.
96–100
Exceptional
Comprehensive, unambiguous, edge cases covered, fully testable. No gaps a reviewer could reasonably surface.
Below 85
Needs Work
Required Clarifications materially compromise clarity or testability. Score impact: Critical = −10 to −20; High = −5 to −10; Medium = −2 to −5.
Score is never penalized for
Absent performance SLAs, accessibility notes, audit trails, notification behaviors, or optional enterprise considerations — unless they are obviously essential for the specific requirement type being reviewed.
Review Standard
What Makes a Requirement Ready for Development?
The review answers one question: can a developer implement this correctly, and can QA verify it objectively? A requirement is ready when it has all four of these elements clearly defined.
Actor
Who initiates or is affected by this requirement?
Trigger
What event or condition causes this behavior?
Outcome
What is the observable result when the trigger fires?
Testable Acceptance Criteria
Can QA write a concrete pass/fail test for each criterion without making assumptions?
See the review engine in action
Sign up and paste a requirement. The review takes a few seconds and you'll see exactly how the five categories and scoring work on your own content.
Get Started