AI Pundit Spec Kit

AI Pundit Spec Kit brings a specification-driven interface directly into Azure DevOps, enabling teams to generate, review and manage structured software specifications, delivery plans, task breakdowns and quality checklists — then automatically create Azure Boards work items from them.

Secure Sign-In

Sign in with your Microsoft account to get started. AI Pundit Spec Kit uses Microsoft authentication to securely connect with your Azure DevOps organization.

Key Features

Conversational AI Chat

Natural-language interface for context-aware specification engineering. Ask questions, give instructions, and let the AI generate structured deliverables — all within Azure DevOps.

Repository-Aware Context

Automatically scans your wiki's file tree and source files to give the AI full project context — README, existing specs, constitution, and project structure.

Wiki-Based Storage & Direct AI Integration

AI Pundit Spec Kit is built on top of spec-kit, an open-source specification engineering framework. spec-kit provides the structured workflow, command system and specification methodology that powers this extension.

All Speckit-generated files (spec.md, plan.md, tasks.md, constitution.md) are automatically saved to your Azure DevOps Wiki with full version history — no Git commits required. The extension makes direct API calls to your chosen AI provider (Azure OpenAI, OpenAI, or Google Gemini) from the browser without any proxy servers or functions, ensuring fast, secure, and cost-effective operations.

Multi-Provider AI Support

Bring your own API key. Supports Azure OpenAI, OpenAI, and Anthropic Claude — configure in Settings.

Settings — Project & Organization Level

Configure AI providers and command behavior from the Settings panel with two levels of control:

• Project Settings — Set up LLM configurations (AI provider, API key, model, endpoint) at the project level. Toggle between Default (inherits organization config) and Custom mode to override with project-specific settings. Customize Commands Mapping to rename or remap any /speckit.* command.
• Organization Settings — Organization admins can define shared LLM configurations (title, AI provider, expiration) that apply across all projects, providing centralized governance and key rotation.

Spec Kit Workflow

AI Pundit Spec Kit follows a structured specification workflow using slash commands:

Define Project Standards (/speckit.constitution)

Establish and maintain your project's constitutional principles. The AI generates a constitution.md that guides all subsequent specifications, ensuring consistency across your project.

Create Specification (/speckit.specify)

Create or update detailed feature specifications (spec.md + requirements.md) with structured sections, acceptance criteria, and a specification quality checklist.

Clarify Requirements (/speckit.clarify)

AI identifies gaps and ambiguities in your specs, generates targeted clarification questions, and integrates the answers directly into the specification.

Create Implementation Plan (/speckit.plan)

Generate comprehensive delivery plans including plan.md, research.md, data-model.md, quickstart.md, and API contracts (OpenAPI/GraphQL).

Generate Implementation Tasks (/speckit.tasks)

Produce phase-based task breakdowns (tasks.md) with task IDs, parallel markers, and user story links — ready to sync with Azure Boards.

Analyze Requirements and Plan (/speckit.analyze)

Analyze specs, plans, and tasks for duplication, ambiguity, underspecification, and coverage gaps. Get severity-rated findings (Critical, High, Medium, Low) with remediation support.

Run Readiness Checklist (/speckit.checklist)

Generate "unit tests for requirements" — domain-specific quality checklists (UX, API, Security, etc.) to validate completeness, clarity, and measurability.

Work Item Integration

Automatically sync your specifications with Azure Boards. The hierarchy chart visualizes your work item structure — Epics, Features, User Stories, and Tasks — with drag-to-pan navigation and zoom controls.

File Management

• Generate files — AI creates structured spec files (spec.md, plan.md, tasks.md, etc.)
• Edit view — Review AI-generated changes before saving
• Keep — Review and keep generated files directly to wiki
• Attach context — Attach existing repo files, wiki pages, or work items to provide additional context

Chat History

Persistent, project-scoped chat sessions let you resume any previous conversation right where you left off. The chat history panel lets you:

• Browse sessions — View all past conversations for your current project
• Resume anytime — Reopen any session to continue where you left off, with full message context replay
• Rename sessions — Give sessions descriptive names for easy navigation
• Delete sessions — Remove individual sessions or clear all history at once
• New chat — Start a fresh conversation without losing prior sessions

Created using Raaghu Design System with AI Pundit leveraging spec-kit.

©2026 Wai Technologies. All rights reserved.