Find & Resolve Drift

Resolving Specs Inconsistencies

Check a feature’s specifications before implementation. Investigate contradictions, duplicates, and unclear behavior, then review the corrections with the team.

Start with Reqode

What This Use Case Covers

Reqode checks whether the product model is coherent enough to implement.

Duplicates

Two requirements, API operations, UI states, or data definitions describe the same behavior in competing ways.

Cross-type contradictions

A requirement, data entity, user interface, wireframe, or API contract says something different about the same rule.

Ambiguity

A requirement leaves open states, actors, edge cases, permissions, validation rules, or error behavior that implementation would otherwise have to guess.

Spec type guidelines

Requirements, Data Entities, User Interfaces, wireframes, and API Operations are checked against the structure expected for their type.

What Module Consistency Verifier Checks

The verifier works with specifications only. It does not inspect source code or software units.

Target Module

The selected requirement module and all requirements and specs inside it.

Data Entities Specs

Data Entities with attributes, enum values, and relations.

UI Specs & Wireframes

User Interfaces, wireframes, and UI transitions.

API Operations

API contracts, request and response fields, errors, and side effects when represented in Reqode.

Cross-type Consistency

Relation maps show whether requirements, data fields, UI actions, transitions, and API contracts support the same product behavior.

Spec Type Guidance

Flag requirement, data, UI, and API specs that do not follow the expected structure, completeness, or clarity for their type.

Examples of Useful Findings

The verifier records concrete spec problems that a team can fix before they become implementation assumptions.

Severity: Medium

Duplicate requirement

Two requirements define the same approval step with different names, statuses, or actors.

References

Requirement: Approval step A
Requirement: Approval step B

Severity: High

Missing data state

A UI transition uses a status that is not present in the linked Data Entity values.

References

User Interface: Status transition
Data Entity: Status values

Severity: High

API mismatch

The requirement expects a rejection reason, but the API contract has no request or response field for it.

References

Requirement: Reject request
API Operation: Reject request

Severity: Medium

Underspecified rule

A requirement says "notify the user" but does not define channel, trigger, recipient, timing, or failure behavior.

References

Requirement: Notify user

Severity: Low

Wrong artifact type

API behavior is buried inside a requirement description instead of being represented in the API Operation spec.

References

Requirement: API behavior

Severity: Medium

Weak type compliance

A Data Entity omits required attribute descriptions, enum values, constraints, or deprecation markers needed for reliable use.

References

Data Entity: Field definitions

Findings Turn Spec Issues into Work

Every Finding carries the evidence. Review the issue, then choose the right path to a resolution.

Finding Illustrative example

Rejection reason is missing from the API

Type: Inconsistency Severity: High

Evidence

Requirement — a rejection must include a reason.

API contract — no rejection reason field is defined.

Affected artifacts

Requirement API Operation

AI Analyst

Investigate the mismatch and prepare specification changes for review.

Proposed corrections → team review

Product Decision

Ask the product owner to clarify which behavior the specifications should describe.

Open question → agreed behavior

Change Request

Group a broader correction across affected artifacts or an isolated branch.

Related changes → coordinated scope

Three ways forward, with the Finding’s evidence and affected specifications as context.

Main Workflow

Use this before implementation, before product or engineering review, or whenever a spec-heavy area changes.

Choose a module

Choose the module for the feature or capability being reviewed.

Run verification

Run Module Consistency Verifier in the selected branch. Its scope is specifications, not source code or software units.

Review Findings

Inspect Finding evidence, severity, affected artifacts, and solution options.

Resolve spec issues

Review corrections to behavior, contracts, type guidance, duplication, or missing relations.

Track broader cleanup

Use a Change Request for changes across artifacts or an isolated branch.

Verify again

Recheck the module and review remaining Findings.

Automated Workflow

Keep spec quality checks running as the product evolves. Set up an AI Automation to check modules that need verification after changes, on your chosen schedule.

Configure the automation

In Project Settings → AI Automations, add a Module Consistency Check. Choose the project branch, AI model, and maximum checks per run.

Set the schedule and enable

Choose a Run Mode, save the rule, and enable it. Each scheduled run picks up modules awaiting verification in the selected branch.

Review and resolve Findings

Review the completed checks and their Findings. Inspect the evidence, correct the specifications, and let scheduled checks follow subsequent changes.

Explore the workflow

Find conflicting dispatch rules in FieldDesk, review and apply the correction, then verify the module again.

Check the dispatch approval rules

Run a module consistency check in the review branch. The two rules disagree on when manager approval is required.

Ready to Gate Spec Quality?

Review one feature’s specification Findings before handing the accepted scope to your coding agent.

Start with Reqode