llm-harness

Fable-style spec + evidence gate for Claude Code + Codex. Makes Opus/Codex work under Fable-like discipline: blocks every edit until a deterministic spec passes, and there is no "done" without live acceptance evidence. Spec-first, verification-gated, forbidden-paths enforced.

Make every model cheaper or better. Zero-config behavioral harness for Claude Code, with a reproducible 408-run benchmark

Persistent memory for AI agents. Extract once, recall forever.

Local AI harness/workstation: local AI image editor, Ollama image generation routing, SDXL inpainting UI, CivitAI model imports, 8GB VRAM Stable Diffusion.

Inspired by GothamChess's AI chess videos - this is an LLM harness for chess. Fixes the real problems that were causing the models to stumble; proper board state context and move validation. Pit GPT, Gemini, Claude and others head-to-head with live reasoning streams, knockout tournaments, and PGN export.

A full-screen terminal environment for writing fiction with language models.

更聚焦普通业务场景的harnass

OpenCode Plugin for supervised agentic software development on a Kanban board

An LLM-native cognitive harness for React + Effect TypeScript projects

METAL-SCI: a scientific compute benchmark for evolutionary LLM kernel search on Apple Silicon Metal

On-prem Claude Agent Skills on a local LLM (LM Studio) — natural-language routing + deterministic executors. Private & reliable. 🇰🇷 HWPX + AI-humanizer included.

Welcome to a daily adventure with Super-Lili.🐝

Per-project AI coding harness for Claude Code · Cursor · Codex. Profiler + 10-dim interview build a different harness for every project.

Model-agnostic autonomous coding agent engine with Claude Code capabilities, subagents, and an 18-target cross-agent SSOT harness

A·R·I is a locally hosted intelligence framework built on one core principle: an AI can only be truly personal if it's truly private.

agentic CLI tool in rust using openrouter

A small code snippet as a part of "How to Write an Agent" Blog Post

smith is a fast, local-first LLM Agent Harness written in Crystal

The general-purpose LLM Harness. Wrap your models with a multi-tenant agent runtime, memory, and strict guardrails.

Shape Up for coding agents — with gates the agent can't talk its way past. A Claude Code plugin that runs idea → pitch → build → eval → ship, with PreToolUse hooks that hard-deny a premature EVAL and progress derived from test artifacts instead of self-report.