v0.8.0 Release Notes¶
Release Date: 2026-06-07
This release introduces Big Tech methodology profiles, constitution spec support for organizational governance, and comprehensive AI assistant integration with agent specs and Claude Code skills.
Highlights¶
- Big Tech methodology profiles with multi-company best practices
- Constitution spec support for organizational governance
- Agent specs and Claude Code skills for AI integration
- MCP resource handlers for programmatic spec access
- Continuous Discovery and Shape Up methodology profiles
What's New¶
Big Tech Methodology Profiles¶
New profiles that combine best practices from leading tech companies:
# Full Big Tech profile (10 methodologies)
visionspec init my-product --profile big-tech-product
# Streamlined Big Tech Essentials (3 companies)
visionspec init my-feature --profile big-tech-essentials-feature
# Scope-specific AWS profiles
visionspec init new-product --profile aws-product
visionspec init new-feature --profile aws-feature
big-tech-product includes:
- Google Design Docs + RFC
- Amazon Working Backwards
- Meta Lightweight Decision Records
- Stripe API-First Design
- Apple Human Interface Guidelines
- Microsoft RAPID Decision Framework
- Netflix Freedom & Responsibility
- Airbnb Design Language System
- Uber API Guidelines
- LinkedIn InVision
big-tech-essentials focuses on three core methodologies: Google, Amazon, and Stripe.
Constitution Spec Support¶
Organizational governance with inheritance hierarchy:
# constitutions/org/acme.yaml
apiVersion: visionspec/v1
kind: Constitution
metadata:
name: acme
level: organization
technical:
languages:
backend:
primary: go
allowed: [go, rust]
tenancy:
model: multi-tenant
isolation: rls
infrastructure:
iac:
tool: pulumi
language: go
Constitutions flow down through inheritance:
org/acme.yaml # Organization defaults
↓ inherits
team/platform.yaml # Team overrides
↓ inherits
project/api.yaml # Project-specific
Claude Code Skills¶
New skills for AI assistant integration:
# Available skills
/visionspec-eval # Evaluate specifications
/visionspec-export # Export to targets
/visionspec-lint # Validate structure
/visionspec-reconcile # Generate spec.md
/visionspec-status # Check readiness
/visionspec-synthesize # Generate specs
Agent Specs¶
Pre-configured agent specifications for:
- Orchestrator - Coordinates spec workflow
- Evaluator - Runs rubric-based evaluation
- Synthesizer - Generates derived specs
- Exporter - Exports to target systems
- Reconciler - Resolves conflicts
MCP Resource Handlers¶
Programmatic access to spec content via MCP:
# Start MCP server
visionspec serve
# Resources exposed:
# - projects/{project}/specs/{type}
# - projects/{project}/status
# - projects/{project}/eval/{type}
Additional Methodology Profiles¶
Continuous Discovery (Teresa Torres):
- Opportunity trees for continuous learning
- Customer interview cadence
- Assumption testing
Shape Up (Basecamp):
- 6-week cycles with pitches
- Appetite-based scoping
- Hill charts for progress
Profile Updates¶
Stage-Based Profiles Enhanced¶
| Profile | New Specs Added |
|---|---|
0-1 |
experiment, lean-canvas |
startup |
trd, uxd |
growth |
trd, tpd, ird |
enterprise |
governance specs |
Legacy AWS Profile Removed¶
The legacy aws profile has been replaced with:
aws-product- New product lines (MRD start)aws-feature- Features on existing products (OpportunitySpec start)
New Packages¶
| Package | Description |
|---|---|
pkg/constitution |
Constitution schema and inheritance |
pkg/apptypes |
Application type definitions |
pkg/version |
Version management utilities |
Documentation¶
- Big Tech and methodology framework guides
- Execution target documentation
- Organization customization guide
- Workflow example documentation
- Agent specs and steering docs
- Complete CLI documentation for all commands (rules, render, watch, version)
Fixed¶
- Windows compatibility: Profile loading now works correctly on Windows by using
path.Joininstead offilepath.Joinfor embedded filesystem paths - MCP resource handlers: Refactored to eliminate code duplication
Installation¶
Requirements¶
- Go 1.22+
- structured-evaluation v0.6.0