Skip to content
| Marketplace
Sign in
Visual Studio Code>AI>AIEXENew to Visual Studio Code? Get it now.
AIEXE

AIEXE

AIEXE

| (0) | Free
AIEXE — Artificial Intelligence Execution Engine for spec-driven development
Installation
Launch VS Code Quick Open (Ctrl+P), paste the following command, and press enter.
Copied to clipboard
More Info

AIEXE

Spec-Driven Development with Claude — VS Code extension that operationalizes the DDS-IA 4.0 methodology for AI-orchestrated software delivery.

Proprietary — internal use. See LICENSE and ADR-0005.

Features

Spec-First Workflow

  • New Slice — Create structured spec documents from templates (10 categories)
  • Inline Validation — Real-time spec quality scoring with CodeLens indicators
  • Plan with Opus — Generate implementation plans using Claude Opus 4.7

Quality Engine

  • Fresh Review — Independent LLM-as-judge review of completed implementations
  • Drift Detection — Background monitoring for CLAUDE.md rule violations
  • CA Compliance — Track acceptance criteria met across slices

Dashboard & Analytics

  • Metrics Dashboard — Cycle time, cost, compliance trends with interactive charts
  • Cost Tracking — Per-operation and per-model cost breakdown
  • Budget Forecast — Monthly burn rate projection

Project Setup

  • Guided Walkthrough — 5-step onboarding for new AIEXE projects
  • Archetype Selection — 8 project archetypes with pre-configured subagent libraries
  • Scaffold Templates — Project initialization with CLAUDE.md, .aiexe/, and structure

Quick Start

  1. Install the extension from VS Code Marketplace
  2. Open a workspace and run Cmd+Shift+P > AIEXE: Start Project Setup
  3. Follow the walkthrough to configure your project archetype and stack
  4. Create your first slice: Cmd+Shift+P > AIEXE: New Slice
  5. View dashboard: Cmd+Shift+P > AIEXE: Show Dashboard

Requirements

  • VS Code 1.85.0 or later
  • Node.js 20.0.0 or later
  • Anthropic API key (for LLM features: planning, fresh review)

Configuration

Setting Default Description
aiexe.freshReview.autoTrigger true Auto-trigger fresh review on slice execution

Commands

Command Description
AIEXE: Show Dashboard Open metrics dashboard
AIEXE: New Slice Create a new slice from template
AIEXE: Plan with Opus Generate implementation plan
AIEXE: Fresh Review with Opus Run independent quality review
AIEXE: Start Project Setup Launch onboarding walkthrough
AIEXE: Check Drift Scan for CLAUDE.md rule violations
AIEXE: Set API Key Configure Anthropic API key
AIEXE: Restart MCP Server Restart the background MCP server

Architecture

AIEXE runs as a VS Code extension with a background MCP server:

  • Extension — UI layer (commands, webviews, CodeLens, TreeViews, status bar)
  • MCP Server — Business logic (SQLite storage, Claude API, spec validation, drift detection)
  • Communication — Model Context Protocol over stdio

Local Development

Prerequisites: Node 20+, pnpm 9+, VS Code 1.85+.

pnpm install
pnpm --filter aiexe-extension build

Press F5 in VS Code to launch Extension Development Host.

Privacy

  • All data stays local (SQLite in .aiexe/state.db)
  • LLM calls go directly to Anthropic API (your API key)
  • No telemetry collected in this version

Publish Setup

To publish, configure these GitHub repository secrets:

  • VSCE_PAT — VS Code Marketplace Personal Access Token
  • OVSX_PAT — Open VSX Registry access token

Then create a GitHub Release to trigger the publish workflow.

License

Proprietary — internal use (ADR-0005).

  • Contact us
  • Jobs
  • Privacy
  • Manage cookies
  • Terms of use
  • Trademarks
© 2026 Microsoft