Show HN: Ordewell – turn one goal into an ordered plan of coding-agent tasks
summary
Ordewell is a multi-agent task orchestration tool that transforms a single coding goal into a structured, editable plan of tasks. It allows users to assign specific AI models and runners to each task, ensuring the most appropriate tool is used for different types of work. The system emphasizes transparency and verification, using a dedicated VerdictEngine to confirm task completion based on evidence rather than AI opinion. By decoupling planning from execution, it provides a flexible workflow for managing complex software development tasks.
key points
- The tool features a read-only planner that researches the codebase and interacts with the user to refine the goal before generating a typed artifact plan.
- It supports a multi-runner architecture where users can integrate various AI agents like Claude Code, Codex, or custom CLI plugins, assigning different models to different tasks.
- Ordewell provides multiple interfaces, including a VS Code extension, a Terminal UI, and a headless CLI, all driven by a a single core API server.
who should read it
Software developers and engineers who use AI coding agents and want a more structured, verifiable way to orchestrate complex, multi-step development goals.
discussion skeptical
consensus
Commenters agree that using a deterministic scaffold or a high-level planner to guide lower-cost LLMs prevents them from getting lost or degrading in quality. There is a general consensus that frontier models are best suited for initial planning and review, while smaller models can handle the execution of specific tasks. However, a significant divide exists regarding the use of AI-generated responses in social interactions, with many viewing it as disrespectful.
top insight
The most valuable perspective is the concept of 'electric motors on each desk' compared to 'central shafts', suggesting that instead of a single chat window, LLM API calls should be integrated into deterministic shell scripts or specialized tools. This approach allows for precise control over execution and reduces the reliance on on-model reliability for the entire workflow. It enables the use of smaller, cheaper models for specific tasks while maintaining overall project stability.
Generated by an LLM from the article and thread. Check the sources before relying on it.