long-runner-harness

What problem does it solve?

Long-Runner Harness provides a structured framework to automate the end-to-end execution of long-running agent workflows, including environment bootstrap, dual-agent coordination, browser-based verification, and secure session isolation.

Core Features & Use Cases

• Browser-first validation: each feature is validated in a real browser via Playwright MCP to ensure UI correctness and visual fidelity.
• Secure session isolation: every feature runs in an isolated context to prevent context bleed between steps.
• Deterministic dispatch: an automated dispatcher drives feature progression with dependency resolution and rollback-safe commits.
• Use cases include autonomously bootstrapping projects, generating incremental features, running browser-based tests, and collecting traces for cross-session learning.

Quick Start

Execute /long-runner init to bootstrap a new harness with your project description.