Documentation
Welcome to the SpecMarket documentation. Learn how to create, run, and publish specs.
Getting Started
Quick start guide for SpecMarket — install the CLI, find a spec, run it, own the result.
Spec Format Reference
Complete reference for the SpecMarket spec directory structure, spec.yaml schema, and all required files.
What is a Ralph Loop?
Understanding the Ralph Loop — the autonomous agentic coding loop at the heart of SpecMarket. How it works, why it works, and what makes a good one.
Security & Trust Model
How SpecMarket keeps you safe — spec scanning, sandboxing, telemetry privacy, community flagging, and what we can't protect you from.
FAQ
Frequently asked questions about SpecMarket, Ralph Loops, specs, costs, security, and getting started.
CLI Command Reference
Complete reference for all 14 SpecMarket CLI commands — authentication, spec lifecycle, registry, collaboration, and configuration.
Publishing Your First Spec
Step-by-step guide to creating, validating, and publishing a spec on SpecMarket — from blank directory to live registry listing.
Contributing to SpecMarket
How to contribute specs, report bugs, improve documentation, and participate in bounties.
Publishing Existing Specs
Publish specs written in Spec Kit, BMAD, Ralph, or any custom format to SpecMarket.
Troubleshooting & Known Issues
Solutions to common problems, known limitations, and workarounds for SpecMarket.
API Reference
Convex queries, mutations, and actions for SpecMarket platform integration.
Managed Runs — Cloud-Hosted Spec Execution
Run specs in the cloud instead of locally. Let SpecMarket handle infrastructure, pay only for what you use.
Advanced Spec Authoring
Patterns, antipatterns, and advanced techniques for writing production-grade specs that scale.
Using SpecMarket with Spec Kit
Spec Kit handles spec authoring. SpecMarket handles execution, verification, and distribution. Here's how to use them together.
Using SpecMarket with BMAD
BMAD structures your multi-agent workflows with role-based orchestration. SpecMarket executes those workflows in a sandboxed environment with full observability. Here's how to use both together.