Show HN: Hanesu – An experimental workflow layer for AI coding agents (github.com)

Hanesu has been introduced as an experimental workflow layer designed to optimize the interaction between AI coding agents and software repositories. Unlike traditional approaches that utilize single prompts for coding tasks, Hanesu establishes a structured workflow incorporating task files, defined phases, and human review gates. This allows AI agents to navigate complex, multi-step tasks with greater predictability and thoroughness, rather than engaging in lengthy and potentially unclear conversations. While Hanesu won't replace existing tools like OpenCode or Codex, it enriches their functionality by enforcing a more strategic coding process. This new framework is significant for the AI/ML community as it emphasizes a more disciplined approach to AI-assisted coding, borrowing principles from Spec-Driven and Test-Driven Development. Key features include a context-first search method, the use of artifacts to document decisions, and risk-based human review gates that ensure quality at critical junctures. Hanesu is particularly useful for ambiguous or high-risk coding scenarios, guiding agents to produce reliable outputs while enabling developers to maintain oversight. By integrating Hanesu's structured workflow directly into coding practices, projects could see improved outcomes in software development through enhanced collaboration between human and machine intelligence.