Documentation

Feature1 Platform Docs

Everything you need to get started with AI-powered SDLC orchestration

Quick Start

1. Connect Repo

Onboard your GitHub, GitLab, or Bitbucket repository in minutes.

2. Build Domain Spec

AI creates a knowledge graph of your codebase, personas, and architecture.

3. Ship Features

Plan features, generate ACs, implement with AI, and create PRs.

CLI Agent Guides

Codex CLI + Claude Code with Feature1 MCP

Connect MCP, authenticate, and run AC-by-AC implementation flows from your preferred coding agent.

Open guide

MCP Integrations Overview

See supported integrations and setup patterns for Feature1 MCP in local and container environments.

Open guide

Platform Overview

What is Feature1?

Feature1 is an AI-powered SDLC orchestration platform that intelligently coordinates every phase of software development — from ideation to deployment. Our orchestration engine maintains context across all development phases, eliminating handoffs and context loss.

Domain Intelligence— Living knowledge graph mapping personas, features, and code
Competitor Analysis— Automated competitive intelligence
Feature Planning— 6-stage AI analysis with auto-generated PRDs
Dual-Mode Implementation— Autopilot (AI executes with approval gates) or Copilot (Driver-Navigator via MCP)
PR Automation— Branch, commit, and pull request creation
Universal Compatibility— Any tech stack, any existing codebase

Core Concepts

Feature Analysis Pipeline

6-stage AI analysis that understands your feature before implementation begins.

Decomposition
Feasibility Assessment
Impact Analysis
Implementation Study
Open Questions
PRD Generation

Acceptance Criteria Engine

Auto-generates testable ACs from user stories, then implements each one individually.

AI-generated from feature analysis
Each AC implemented separately
Validation after each implementation
Criterion-by-criterion approval in Copilot mode

Domain Spec

A living knowledge graph that maps your product domain and evolves with your codebase.

Persona mapping
Feature-to-code relationships
Architecture understanding
Auto-regenerates as codebase changes

Sprint Management

Plan sprints, manage backlogs, and track velocity — integrated with the implementation pipeline.

Sprint creation with goals and dates
Backlog assignment and prioritization
Velocity tracking across sprints
Full audit trail

Feature1 MCP Server

Feature1 exposes 30+ tools via the Model Context Protocol (MCP), allowing any MCP-compatible AI agent to interact with Feature1 workflows.

Claude Code + Feature1 MCP
$ claude --mcp feature1
> login(email, password)
✓ Authenticated
> load_workflow("add-stripe-payments")
→ Workflow loaded: 6 acceptance criteria
> get_next_ac()
→ AC001: Payment form renders with card input
> implement_ac(criterion_index: 0)
✓ Implementation complete, tests passing
> mark_implementation_done(criterion_index: 0)
✓ AC001 approved — moving to AC002

Available MCP Tools

login
load_workflow
get_workflow_state
get_next_ac
generate_ac_prompt
implement_ac
mark_implementation_done
approve_criterion
create_pr
create_feature
create_user_story_draft
get_feature
list_features
list_user_stories
get_sprint_summary
get_velocity_summary
get_codebase_context
get_repository_context
search_platform_artifacts
find_related_features
get_planning_context

Compatible Agents

Claude Code
Cursor
Windsurf
Any MCP Agent

Execution Modes

Autopilot

AI implements each acceptance criterion autonomously — commits code and opens PRs — pausing at approval gates for you to confirm "Looks good" or "Needs fix."

Best for: routine features, bug fixes, boilerplate

Copilot

Follows the Driver-Navigator pair programming pattern. The MCP client (Claude Code or Codex CLI) drives the implementation while you navigate — guiding design decisions and architecture choices at each step.

Best for: critical features, new patterns, complex logic

Seat-Based Pricing

Seat TypePriceIncludes
PM Seat$300/moFeature planning, analysis, sprint management, F1 Assistant
Designer Seat$150/moUI/UX workflows, Figma MCP integration
Dev Seat$20/moCode implementation, PR reviews, MCP tools
QA Seat$20/moTest recording, bug capture, AC validation
Autopilot License $1,000/moHigh-automation implementation with HITL approval gates

Integrations

GitHub
GitLab
Bitbucket
Azure DevOps
Figma
Claude Code
Cursor
Windsurf

FAQ

Do you replace developers?

No, we augment them. Copilot mode keeps developers in control while Autopilot handles routine work.

What about code quality?

We enforce SOLID principles, clean code practices, and comprehensive testing. Zero technical debt from AI-generated code.

How is this different from GitHub Copilot?

Copilot suggests code snippets. Feature1 plans, builds, tests, and deploys complete features end-to-end with full SDLC orchestration.

What tech stacks do you support?

Any. React, Vue, Python, Node.js, Go, Java, PHP, .NET — Feature1 works with any programming language and framework.

Can I use my own AI coding agent?

Yes. Feature1 MCP server works with Claude Code, Cursor, Windsurf, and any MCP-compatible agent.

Ready to get started?

Connect your repository and start shipping features with AI.

Book Demo