Skip to content

Choosing a Target

This guide helps you select the right execution target for your project.

Quick Decision Matrix

If you need... Choose
Enterprise rigor with approval gates AWS AI-DLC
GitHub-native PR workflows SpecKit
Fast iteration with parallel execution GSD
Complex DAG-based task orchestration GasTown
Role-based multi-agent coordination GasCity

Feature Comparison

Feature AI-DLC SpecKit GSD GasTown GasCity
Execution Model 3-phase lifecycle Sequential Wave-parallel DAG Role-based
Parallel Execution Per-phase No Yes (waves) Yes (DAG) Yes (roles)
Multi-Agent Specialists No No Yes (beads) Yes (agents)
Approval Gates Yes Implicit (PRs) No No Optional
Audit Trail Complete Git history STATE.md Formula state City state
Human-in-Loop Required PR review Optional Optional Optional
GitHub Integration External Native External External External

By Project Type

Startup MVP

Recommended: GSD

Speed > Ceremony
Parallel execution
Fast iteration
Minimal overhead
visionspec init my-mvp --profile startup
visionspec export gsd

Enterprise Feature

Recommended: AWS AI-DLC

Compliance required
Audit trails
Approval gates
Multi-stakeholder
visionspec init my-feature --profile enterprise
visionspec export aidlc

Open Source Project

Recommended: SpecKit

GitHub-native
PR-based review
Sequential tasks
Community contributors
visionspec init my-oss --profile startup
visionspec export speckit

Large-Scale System

Recommended: GasTown or GasCity

Complex dependencies
Multiple agents
Parallelizable work
Specialized roles
visionspec init my-platform --profile enterprise
visionspec export gastown  # or gascity

By Team Structure

Solo Developer

Scenario Target
Quick prototype GSD
Careful implementation SpecKit
Complex project GasTown

Small Team (2-5)

Scenario Target
Collaborative PRs SpecKit
Parallel workstreams GSD
Shared agents GasTown

Large Team (5+)

Scenario Target
Role specialization GasCity
Compliance needs AI-DLC
Mixed complexity GasTown

Enterprise

Scenario Target
Regulated industry AI-DLC
Multi-team coordination GasCity
Complex orchestration GasTown + GasCity

By Methodology Pairing

Different VisionSpec frameworks pair naturally with different targets:

AWS Working Backwards + AI-DLC

Natural pairing
Same philosophy
Enterprise-grade
Full lifecycle
visionspec init project --profile enterprise
# Use AWS framework
visionspec export aidlc

Lean Startup + GSD

Build-measure-learn
Fast iteration
Minimal ceremony
Quick pivots
visionspec init experiment --profile startup
# Use Lean Startup framework
visionspec export gsd

Google Design Docs + SpecKit

RFC-style specs
Sequential implementation
PR-based review
Design doc tradition
visionspec init design --profile growth
# Use Google framework
visionspec export speckit

Stripe API-First + GasCity

API contracts first
Role specialization
Backend/Frontend split
Clear boundaries
visionspec init api --profile enterprise
# Use Stripe framework
visionspec export gascity

Complexity Analysis

Simple Projects

Characteristics:

  • Single component
  • < 10 tasks
  • No parallelization needed
  • Single developer

Best Targets: SpecKit, GSD

Medium Projects

Characteristics:

  • Multiple components
  • 10-50 tasks
  • Some parallelization
  • Small team

Best Targets: GSD, GasTown

Complex Projects

Characteristics:

  • Many components
  • 50+ tasks
  • High parallelization
  • Large team
  • Compliance needs

Best Targets: GasTown, GasCity, AI-DLC

Migration Paths

Starting Simple, Growing Complex

Start: GSD
  ↓ (need structure)
Migrate: SpecKit
  ↓ (need parallelization)
Migrate: GasTown
  ↓ (need roles)
Migrate: GasCity

Starting Enterprise

Start: AI-DLC
  ↓ (need multi-agent)
Add: GasTown/GasCity for implementation
Keep: AI-DLC for governance

Hybrid Approaches

AI-DLC + GasCity

Use AI-DLC for lifecycle management, GasCity for implementation:

# Inception with AI-DLC
visionspec export aidlc

# Construction with GasCity
visionspec export gascity

GasTown + GasCity

Use GasTown for task DAG, GasCity for agent coordination:

# Task breakdown
visionspec export gastown

# Agent assignment
visionspec export gascity

Cost Considerations

Target Token Usage Agent Count Typical Cost
GSD Low 1 $
SpecKit Low 1 $
AI-DLC Medium 1-3 $$
GasTown High 3-5 $$$
GasCity High 3-10 $$$$

Decision Flowchart

Start
Need compliance/audit?
  ├─ Yes → AI-DLC
  No
Need multi-agent?
  ├─ Yes → Need role specialization?
  │          │
  │          ├─ Yes → GasCity
  │          │
  │          └─ No → GasTown
  No
Need parallel execution?
  ├─ Yes → GSD
  No
GitHub-native workflow?
  ├─ Yes → SpecKit
  No
Default: GSD (simplest)

Summary

Target Best For Avoid When
AI-DLC Enterprise, compliance Speed is critical
SpecKit GitHub projects, PRs Need parallelization
GSD MVPs, fast iteration Complex dependencies
GasTown Complex DAGs Simple projects
GasCity Role-based teams Solo development

Next Steps