Skip to content

Awesome list for AI agent harness engineering: tools, patterns, evals, memory, MCP, permissions, observability, and orchestration.

Add Continual Harness to Generators & Meta-Harnesses section

ai-boostf094347d809d
diff --git a/README.md b/README.md
index bab0641..b7cc97d 100644
--- a/README.md
+++ b/README.md
@@ -387,6 +387,7 @@ Real repositories worth studying — each with a note on *why* it's worth your t
- [stanford-iris-lab/meta-harness](https://github.com/stanford-iris-lab/meta-harness) — April 2026 official implementation of the Meta-Harness paper from Stanford's IRIS Lab. Provides the framework and two reference experiments for end-to-end harness optimization via filesystem-backed search loops where a coding agent proposes, evaluates, and refines harness artifacts. The cleaned-up codebase is the definitive starting point for researchers reproducing or extending meta-harness optimization. ![Stars](https://img.shields.io/github/stars/stanford-iris-lab/meta-harness?style=flat-square&label=★&color=yellow)
- [autocontext](https://github.com/greyhaven-ai/autocontext) — Recursive self-improving harness that runs multi-generation evaluation loops, distilling successful strategies into persistent playbooks and trace datasets that future agents inherit. The five-role architecture (competitor, analyst, coach, architect, curator) and built-in production trace capture make it the most complete open-source reference for turning meta-harness optimization into a deployable system with MCP support. ![Stars](https://img.shields.io/github/stars/greyhaven-ai/autocontext?style=flat-square&label=★&color=yellow)
- [retro-harness](https://github.com/wbopan/retro-harness) — Official implementation of RHO (Retrospective Harness Optimization): improves an agent's harness using only its own past trajectories, via self-validation, self-consistency, and pairwise self-preference — no ground-truth labels or external evaluators required. A single round moves SWE-Bench Pro from 59% to 78%, showing that harness evolution can be fully self-supervised rather than dependent on labeled validation data. ![Stars](https://img.shields.io/github/stars/wbopan/retro-harness?style=flat-square&label=★&color=yellow)
+- [Continual Harness](https://github.com/sethkarten/continual-harness) — Reset-free framework that lets an LLM Refiner rewrite its own system prompt, sub-agents, skills, and memory mid-episode via an `evolve_harness` tool. The first open-source implementation of online harness self-improvement with reproducible long-horizon benchmarks (Gemini Plays Pokémon). ![Stars](https://img.shields.io/github/stars/sethkarten/continual-harness?style=flat-square&label=★&color=yellow)
- [Omnigent](https://github.com/omnigent-ai/omnigent) — Databricks' open-source meta-harness (June 2026) that sits above Claude Code, Codex, Pi, and custom agents to compose, govern, and share live agent sessions from one control plane. The key harness insight is that enterprises don't need another agent framework — they need a portability and governance layer above the agents they already run, with policy-driven sandboxing and cost caps enforced outside any single tool. ![Stars](https://img.shields.io/github/stars/omnigent-ai/omnigent?style=flat-square&label=★&color=yellow)
### Demo Harnesses