crshdn/mission-control

What it solves

Autensa solves the manual burden of product management and software development by automating the entire lifecycle from market research and feature ideation to coding and pull request creation. It prevents the "blank page" problem for developers by continuously generating data-backed feature ideas and executing them autonomously.

How it works

Autensa operates as an autonomous product engine that integrates with a codebase and a live URL. It follows a structured pipeline:

1. Research & Ideation: AI agents analyze competitors, SEO, and user intent to generate scored feature ideas.
2. Human-in-the-loop Decision: Users review these ideas via a Tinder-style "swipe" interface (Pass, Maybe, Yes, Now!).
3. Planning & Execution: Approved ideas are turned into detailed specs. A multi-agent pipeline (Builder $\rightarrow$ Tester $\rightarrow$ Reviewer) implements the code in isolated workspaces using git worktrees.
4. Delivery: The system automatically creates a GitHub Pull Request after the code passes internal AI review and testing.

Who it’s for

It is designed for developers, product managers, and founders of MVPs or side projects who want to automate the iterative improvement of their software products.

Highlights

• Product Autopilot: A continuous loop of autonomous research, ideation, and implementation.
• Convoy Mode: Parallel execution of large features decomposed into a dependency graph (DAG) with auto-recovery for stalled agents.
• Preference Learning: A per-product model that learns a user's taste based on their swipe history.
• Operator Chat: Real-time communication with agents mid-build via queued notes or direct messages.
• Workspace Isolation: Uses git worktrees and dedicated port allocation to prevent conflicts during concurrent builds.
• Automation Tiers: Flexible control levels ranging from Supervised (manual merge) to Full Auto (end-to-end deployment).