Turn Claude Code into its own Meta-Harness — a skill that evolves the scaffolding around a fixed model (memory, retrieval, context, prompts) via a native propose→score→Pareto loop. Native reimplementation of Meta-Harness (Lee et al. 2026).

Automated harness evolution for AI agents. A Claude Code plugin that iteratively optimizes system prompts, routing, retrieval, and orchestration code using full-trace counterfactual diagnosis. Based on Meta-Harness (Lee et al., 2026).

Meta-harness optimization loop wired onto Islo sandboxes. POC: 0/5→5/5 in four proposer steps. Built on islo.dev.

Local harness substrate for project-scoped agent sessions.

Open-source search engine that automatically evolves your AI agent's prompts, tools, and harness logic through iterative evaluation — plug in Claude Code, Codex, Hermes, or any CLI agent.

A reasoning harness for human-AI collaboration. Preserves what's understood, not just what's said. Open-source skill for Claude Projects. Patent pending.

Project page for Meta-harness on Islo (POC). https://zozo123.github.io/meta-harness-on-islo-page/

작업 해석·도메인팩·에이전트 바인딩·검증 계약을 정의하는 문서형 메타 하네스 | Documented meta-harness for task interpretation, domain packs, agents, and validation

Forge Engineering — Meta-engineering discipline above Harness Engineering. Agents that rewrite their own harness with versioned, benchmark-gated control.

A meta-harness playground template for all your AI agents. Switch seamlessly between GitHub Copilot SDK, Claude Agent SDK, Antigravity SDK, Codex SDK, and Cursor SDK. Build your own playground by extending the template with your tools, skills, MCPs, CLIs, and containerized environments for portable, reproducible agent workflows.

Preserve reasoning chains and logical state in Claude Projects to maintain context and build project understanding across AI sessions.