Documentation
Everything you need to know about Trinity — from your first project to advanced configuration.
Get startedGetting Started
Welcome
Core concepts — Projects, PRDs, Stories, and the Agent Pipeline.
Prerequisites
Git, GitHub CLI, Node.js, Claude Code, and Codex — what to install and how to configure it.
First Project
Step-by-step guide from launch to your first shipped PR.
Navigating
UI layout, sidebar navigation, project switcher, and keyboard shortcuts.
Project Setup
New Project Setup
Set up a greenfield project by talking to the Architect — clarify, design, stack, setup, and skills, then generate your roadmap.
Import Existing
8-step import flow: Scan, Files, Repos, Design, Docs, Skills, Save, Done.
Project Settings
Branching, merge strategy, execution gates, Docker, and per-repo overrides.
Project Assets
Upload wireframes, specs, and brand guides with AI-generated descriptions.
Planning
Dashboard
PRD management, progress tracking, phase overview, and blocked story visibility.
PRD Planning
5-phase AI pipeline: Architect, Story Writer, Dependency Mapper, Package Mapper, Calibrator.
Architect
The planning workspace — drafts, workspace slots, and the path from a request to a committed plan.
Dependency Graph
Visual dependency graph with layout management and cross-PRD views.
Execution
Stories
Story structure, status lifecycle, acceptance criteria, tags, and linking.
Running
Workers, the 4-agent pipeline, git worktrees, monitoring, and retries.
Gates
Every checkpoint a run can pause at — deviation approval, missing secrets, PR review, merge conflicts, and more.
Releases
Group PRDs into shippable units with semver tagging, dependency DAGs, and release branches.
Quality Checkpoints
Quality checkpoints that audit multiple stories before release.
Insights
Account & Workspaces
Signing In
Device code authentication — sign in from the desktop app via trinityailabs.com.
Workspaces
Create workspaces, invite members, assign roles, and collaborate on shared projects.
Inbox
One surface for everything waiting on you across every workspace — invitations, requests to answer, and notifications.
Configuration
App Settings
API keys, default behaviors, execution preferences, and theme selection.
AI Models
Four-tier model system across ten providers, with per-model reasoning effort levels.
Secrets
Encrypted store for every API key, token, and config file your project needs.
Runtime (Skills, Commands & Hooks)
Manage the skills, commands, and hooks your project's agents use, and review AI-authored or imported drafts before they go live.