Open-Source Coding Harness: Building Like Claude Code
Roger_M_Taylor · x · 2026-07-20
Developer @akshaypachaar built a fully open-source coding harness similar to Claude Code. A harness acts as a code wrapper around the LLM; while the model only decides the next step, the harness manages planning, tool usage, memory, and security. Key components include: a core agent loop, file tools (acting as external memory), long-term task planning, sub-agents (with independent contexts), sandboxed execution (ephemeral VMs), human-in-the-loop approvals, persistent memory, and checkpoints. It is built on the CrewAI framework.
Related event: Developer Releases Open-Source Coding Harness(2 posts)→
More from coding & agent
- Spring AI 2.1.0-RC1 adds tool-call confirmation, pausing agent loops before real-data access — therealdanvega · 2026-10-26
- Morgan Stanley and US banks block third-party AI agents over credentials — thione · 2026-10-26
- METR red-teams OpenAI's agent monitoring, finds bypasses and gaps — thione · 2026-10-26
- OpenAI: coding agents weakened safety monitoring, one lied about it — thione · 2026-10-26
- ConvFill: A 0.6B Talker Buys Time While a Large Reasoner Whispers the Real Answer — andimarafioti · 2026-10-26
- 250 Builders Hack With Codex at Berlin Community Hackathon Ahead of OpenAI DevDay — paw_lean · 2026-10-26