Loop Engineering — a reversible, observable multi-agent Claude Code harness for brownfield codebases. Policy lives in exit codes and artifacts on disk, not in prompts.

Evidence-driven development loops for agy, Codex, Claude Code, and opencode

Model-facing fresh-agent Ralph loop over the workflow and subagent seams

System prompt assembly registry for the DeepSeek Harness

Agent-plane presentation selector: composes one agent's tools as Code Mode, native, or both

The concrete agent loop plugin for the DeepSeek Harness

Freeze the model, evolve the harness. Two measured applications: (1) SWE-bench code-repair — conformant GLM->Opus empty-patch cascade resolves 51.3% Lite (n=300) and 55.6% Verified (278/500, Wilson 95% CI [51.2, 59.9], official swebench gold eval, no gold

A verifiable self-improvement loop for agent harnesses. Freeze the model. Evolve the harness. Promote only what proves lift.

reddb.io dev plugin - engineering skills for code agents (autonomous /afk loop, /go dispatch, triage, tdd, diagnose, graph-aware codebase understanding, ...)

Curated repo-health signals (gatherSignals + registry) as a shared leaf package

Scaffold loop engineering patterns and starters into any project. Optional --with-foundry harness stack. Supports Grok, Claude Code, Codex, Opencode and more. npx @cobusgreyling/loop-init . --pattern daily-triage --tool opencode --with-foundry

Token-meter-driven compaction policy and LLM summarization backend for the DeepSeek Harness

A visual workspace that helps AI coding agents plan, build, review, and ship reliable software.

worker-thread workflow engine: executes model-written orchestration scripts off the host event loop, bridging agent() calls back to ctx.subagents

omcc — oh-my-customcode 2.0. Claude-Code-native standalone terminal harness (TUI, DuckDB memory, lesson-learning). Name reserved; full release in development.

CLI for Harness Engineering toolkit

Open-source coding-agent harness you can actually change — own the loop (prompts, gates, routing, skills, terminal states), use any model, run long tasks while you're away.

Reverse Engineering Unicorn — composite skill for Claude Code: reverse-engineer any company into an actionable launch playbook (+ clickable CJM prototype) via a 6-module pipeline with QUICK/DEEP/VERIFIED modes

Event-sourced session store for the DeepSeek Harness

Tool-result spill policy for the DeepSeek Harness — replaces oversized plain-text tool results with a retained preview plus a spill-file path (no service surface)