Ship production-ready code
with AI-guided specs
Plan new features, productionalize prototypes, and hand off agent-ready task prompts to your coding agents. Local-first by default; cloud optional with explicit consent.
Plan with AI-guided clarification, PRDs, and tech specs
Productionalize prototypes with one command
Deliver agent-ready task prompts for coding agents
Capabilities
Everything you need to ship safely
Plan, productionalize, and execute with security, compliance, and agent-ready task prompts baked in.
Planning Workflow
Interactive AI-guided clarification, PRDs, and tech specs with human-in-the-loop approvals.
Productionalize
One command to move from prototype to production readiness across security, SOC 2, and code quality.
Agent-Ready Tasks
Generated prompts designed to hand off directly to your coding agents.
Local-First, Cloud-Optional
Run locally with Ollama or opt into cloud models. Keys stay in your OS keychain.
Semantic Code Intelligence
Tree-sitter AST parsing plus hybrid vector search for deep code understanding.
SAST Integration
Optional Semgrep, Gitleaks, and Trivy scans to catch vulnerabilities early.
How it works
Plan → Productionalize → Execute
Three commands to get from idea to production-ready code, with agent-ready task prompts as the handoff.
Plan
ship-spec planning "Add OAuth authentication" --cloud-ok
AI-guided clarification → PRD → Tech Spec → agent-ready tasks. Human-in-the-loop approvals at every stage.
Productionalize
ship-spec productionalize "B2B SaaS handling PII" --enable-scans --cloud-ok
One command for security, SOC 2, code quality, dependencies, testing, and configuration. Optional Semgrep, Gitleaks, and Trivy scans.
Execute
Tasks → coding agent
Copy the generated task prompts into your coding agent (Cursor, Copilot, etc.) and ship with confidence.
Outputs include PRDs, tech specs, readiness reports, and agent-ready task prompts saved under .ship-spec/ for easy handoff.
See it in action
Real commands, production-ready outputs
Move from idea to production readiness with a handful of commands. Every run produces agent-ready task prompts so your coding agents can execute immediately.
Plan a feature
Clarification → PRD → Tech Spec → Tasks
Production-ready analysis
Security, SOC 2, code quality, dependencies, testing, configuration
Local-only mode
Run entirely locally with Ollama; no cloud data leaves your machine
Why Ship Spec
Built for spec-driven, agent-ready delivery
Compare Ship Spec to manual reviews, traditional SAST, and generic AI assistants.
| Category | Manual review | Traditional SAST | Generic AI | Ship Spec |
|---|---|---|---|---|
| Manual reviews | Varies by reviewer, limited coverage | N/A | No codebase context | Consistent, RAG-backed, full codebase coverage |
| Compliance | Requires deep expertise | Limited mapping | Generic advice only | SOC 2, OWASP ASVS, NIST SSDF, SRE mapping |
| Output | Notes and comments | Vulnerability lists | Text responses | Agent-ready task prompts + PRDs + tech specs |
| Execution model | Human-driven only | Rule-based scans | Freeform chat | Structured workflows + task prompts for agents |