google-research/harness2
Can I trust it?
No red flags in our checks.
Stars from real accounts OK
Our fake-star check passed. In a sample of 182 stargazers, 1% were accounts under a month old and 3% had no followers or repos.
Active OK
Last change 6 days ago.
Licence OK
Apache-2.0: you can use it for free, also in a commercial product. Keep the licence notice.
Quick checks from public data. They help, but they are no guarantee.
What it is
A Python framework for recursive harnessing of AI agents, which optimizes the prompts, tools, memory, skills and orchestration around a fixed model. It refines the harness per task through propose, probe and compose steps and keeps a persistent improvement harness across tasks…
Who it's for
Researchers and developers evaluating or tuning agent harnesses on professional benchmarks with fixed model weights
How to use it
Ask your AI to set it up
Paste this into Claude, ChatGPT or your coding assistant.
Help me install and set up google-research/harness2 in my project. Follow its README: https://github.com/google-research/harness2. Explain each step and check it is safe before running anything.
Star curve
On the radar
Its place among the radar's top 100, day by day, with #1 at the top. Gaps are days it was out of the top 100.
Records
- #1 on RepoAlerts on 19 Oct 2026
- In the top 10 for 2 days, first on 18 Oct 2026
- In the top 100 for 2 days since 18 Oct 2026
On Hacker News
- Harness2 by Google Research – Recursive Agent Harnessing77 points · 18 comments · 19 Oct 2026
Get alerts for repos like this
We'll email you when something in AI agents starts rising. Free, once a week, only what matches.
18,267 GitHub repos watched and ranked every hour. Figures as of 20 October 2026.
What arrives: one email on Mondays with the repos that match, never more. See last week's. Privacy
For maintainers
Add this badge to your README. It shows the repo's place on the radar, updates every hour and links back to this page.
[](https://repoalerts.com/repo/google-research/harness2?ref=badge)Similar repos
- IsuruMaduranga/harness-engineering-101
A book-length series in Markdown that explains how AI agents work by building a roughly 300-line toy agent harness in plain Python. It covers the LLM message array, tool calls, the agent loop, context
- zgcagi/Harness-RSI
A survey and curated reading list on harness self-evolution, where an LLM agent's persistent harness state (prompts, memory, skills, tools, and control logic) is updated from execution experience whil
- AgnesAI-Labs/agnes-harness
A pluggable TypeScript agent harness for Forward Deployed Engineering. It connects models, tools, task state and business interfaces through plugins, runs and records agent work with approvals and a c
- Gen-Verse/DiscoveryRSI
A Python harness for recursive, hypothesis-centred research workflows in life science. It runs multi-agent hypothesis development, records revision lineage and evidence, compiles selected investigatio
- AI4AI-for-Anything/ResearchBoost-Agent
An autonomous agent that profiles bottlenecks in AI training code, implements optimizations, and validates speedups against a frozen judge that checks bitwise equivalence and statistical training equi
- wangmiaozero/pi-harness
A desktop workspace built on Pi Coding Agent for macOS, Windows and Linux. It adds run tracing, replay, budgets, tool policies, checkpoints, recovery, evaluation, and multi-agent task orchestration wi