Your AI product manager

Your AI product manager that helps your team build

LFG Agent is your AI product manager. It analyzes requirements, creates PRDs and technical plans, generates prioritized tickets, and orchestrates Claude Code sessions to build your product. You stay in control.

Try LFG Agent See Features

PRDs

Auto-generated

Tickets

Created & prioritized

Code

Sandboxed sessions

Ship

Review & merge

See it in action

Your AI product manager at work

From requirements analysis to ticket execution, every step is visible and trackable.

PRD generation

Analyze requirements. Generate PRDs.

Feed your idea into the AI chat. LFG Agent reads through your context-heavy data, analyzes requirements, and generates a comprehensive product requirements document with feature breakdowns, technical approach, and stack recommendations.

  • Full PRD with features, priorities, and acceptance criteria
  • Technical analysis with stack and architecture decisions
  • Document versioning — iterate and refine with the agent
LFG Agent — PRD & Analysis
LFG Agent generating a PRD with feature breakdown and technical analysis
LFG Agent — Ticket Board
LFG ticket board with Open, In Progress, Review, and Done columns
Ticket management

Create and prioritize tickets automatically.

Once the PRD is approved, LFG Agent breaks it into user stories and implementation tickets — scoped, prioritized, and ready for execution. All within the app. No external tools needed.

  • User stories with acceptance criteria auto-generated
  • Kanban board with Open, In Progress, Review, Done
  • Priority and dependency-aware ordering
Agent execution

Track and course-correct Claude Code sessions.

Each ticket is executed by Claude Code in a sandboxed environment. LFG Agent tracks progress, shows you exactly what was built, and lets you continue, restart, or course-correct at any point.

  • Sandboxed Claude Code sessions per ticket
  • Live execution logs, git diffs, and server preview
  • Continue or restart sessions with full context
LFG Agent — Ticket Execution
LFG ticket execution showing Claude Code session with completed implementation

What LFG Agent does

Your AI product manager, end to end

LFG Agent handles the entire product lifecycle — from understanding what needs to be built to making sure it gets built right.

Requirements Analysis

Reads through your context-heavy data — briefs, docs, conversations — and distills it into clear, structured requirements. Asks the right questions so nothing gets missed.

PRD Generation

Generates comprehensive product requirement documents with feature tables, priority levels, technical approach, and stack recommendations. Iterate and version them with the agent.

Technical Planning

Selects the right stack, designs system architecture, maps data models and APIs. Produces a technical plan that becomes the blueprint for implementation tickets.

Ticket Creation

Breaks the plan into scoped, dependency-aware tickets with user stories and acceptance criteria. All created within the app — no external project management tool needed.

Claude Code Sessions

Each ticket is handed off to Claude Code running in a sandboxed environment. Track execution logs, view git diffs, preview the running server, and course-correct when needed.

Versioning & Iteration

PRDs and technical documents are versioned as you iterate. Refine scope, adjust priorities, and re-plan — the agent keeps full context across every revision.

The pipeline

From idea to working product

You describe what you need. LFG Agent handles every step. You review and approve.

01 — You

Describe your idea

Share what you need in plain English. A product concept, a feature set, a whole system.

02 — Agent

PRD & plan

Agent generates a full PRD, technical architecture, and implementation plan you can review and iterate on.

03 — Agent

Tickets created

The plan is broken into prioritized tickets with user stories. Each is scoped and queued for execution.

04 — Agent

Code & test

Claude Code executes each ticket in a sandbox. You can watch, course-correct, or let it run autonomously.

05 — You

Review & ship

Review the working product. Approve and merge, or leave feedback. The agent iterates instantly.

You touch 2 of 5 steps. The rest is fully autonomous.

Agents ready

Tell us what to build

Describe your development needs. We'll send you a PRD + plan — and we might build it for free.

Open source. Self-host it.

LFG is open source. Clone it, run it, own it.

bash — lfg 
# Clone the repository
git clone https://github.com/lfg-hq/lfg.git

# Install dependencies
pip install -r requirements.txt

# Run migrations
python manage.py migrate

# Start server
uvicorn LFG.asgi:application --host 0.0.0.0 --port 8000