Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
A runtime substrate that turns an agent's execution into a reversible, Git-like trace, so meta-agents can observe, fork, replay, and revert any run. Couples agent and environments in a copy-on-write fork ~5x faster than docker commit, with ~95% KV-cache reuse on replay. Framework built for meta-agents to supervise, optimize, and train other agents
| Date | Stars |
|---|---|
| 2026-07-24 | 1549 |
| 2026-07-25 | 1554 |
| 2026-07-28 | 1577 |
| 2026-07-30 | 1598 |
| 2026-07-31 | 1601 |
| 2026-08-06 | 1656 |
| 2026-08-15 | 2306 |
| 2026-08-18 | 2319 |
| 2026-08-19 | 2329 |
| 2026-08-20 | 2352 |
| 2026-08-21 | 2359 |
| 2026-08-22 | 2362 |
| 2026-08-23 | 2365 |
| 2026-08-24 | 2370 |
| 2026-08-25 | 2372 |
| 2026-08-26 | 2374 |
| 2026-08-27 | 2376 |
| 2026-08-29 | 2378 |
| 2026-08-30 | 2379 |
| 2026-08-31 | 2382 |
| 2026-09-01 | 2388 |
| 2026-09-02 | 2392 |
| 2026-09-03 | 2394 |
| 2026-09-04 | 2396 |
| 2026-09-05 | 2397 |
| 2026-09-06 | 2396 |
| 2026-09-07 | 2398 |
| 2026-09-08 | 2399 |
| 2026-09-09 | 2401 |
| 2026-09-10 | 2402 |
| 2026-09-11 | 2405 |
| 2026-09-12 | 2407 |
| 2026-09-13 | 2408 |
| 2026-09-14 | 2410 |
| 2026-09-16 | 2414 |
| 2026-09-17 | 2426 |
| 2026-09-18 | 2430 |
| 2026-09-19 | 2431 |
| 2026-09-20 | 2433 |
Today
+2 stars today
This week
+25 stars this week
This month
+74 stars this month
Momentum
68.9
growth rate 1.04%/day
<div align="center">
<img src="https://shepherd-agents.ai/assets/logo-shepherd.png" alt="Shepherd" width="140">
<h1>Shepherd: Programmable Meta-Agents via Reversible Execution Traces</h1>
 [](https://pypi.org/project/shepherd-ai/) [](https://pypi.org/project/shepherd-ai/) [](https://shepherd-agents.ai/) [](https://docs.shepherd-agents.ai/) [](https://arxiv.org/abs/2605.10913) [](https://shepherd-agents.ai/blog)
</div>
---
> [!IMPORTANT]
> **Shepherd is in early alpha** and under active development.
> APIs may still change between releases. Feedback and issues are very welcome!
<p align="center">
<a href="#installation">Install</a> |
<a href="#quickstart">Quickstart</a> |
<a href="#permissions-the-signature-is-the-permission-surface">Permissions</a> |
<a href="#examples">Examples</a> |
<a href="https://docs.shepherd-agents.ai/">Docs</a> |
<a href="#citation">Citation</a>
</p>
**Shepherd** is a runtime substrate for agent work that needs inspection,
reversibility, and supervision. It records agent runs as durable, inspectable
execution traces, with retained workspace outputs that can be reviewed before
they are selected, applied, released, or discarded.
> **Platforms.** Shepherd requires **Python 3.11+**. OS-level grant enforcement
> is executed on **both macOS** (Seatbelt) and **Linux** (Landlock, in a privileged
> container). **Windows is unsupported** (enforcement would be
> advisory-only at best) — use **WSL**.
## Installation
```bash
pip install shepherd-ai
```
Working on Shepherd itself? Install the local editable closure instead:
`python -m venv .venv && . .venv/bin/activate && pip install -r requirements-dev.txt`
(see [CONTRIBUTING.md](https://github.com/shepherd-agents/shepherd/blob/main/CONTRIBUTING.md)).
## Quickstart
Shepherd is an agent framework: a task's implementation can be a sandboxed
agent, and its work comes back as a **reviewable proposal** — nothing touches
your files until you accept it. Here the whole body of a task *is* a Claude
agent.
> Needs the `claude` CLI — signed in (a Claude subscription works) or with an
> `ANTHROPIC_API_KEY`. Neither? Jump to the
> [Offline Quickstart](#offline-quickstart) — it runs anywhere, keyless.
>
> On a subscription, a sandboxed run is most reliable with a long-lived token:
> `export CLAUDE_CODE_OAUTH_TOKEN=$(claude setup-token)`. A short-lived signed-in
> session can't be refreshed from inside the sandbox, so it may work
> interactively yet fail here — `shepherd doctor claude` (add `--probe` for a real
> auth round-trip under Shepherd's config, in the parent — not a jailed run) tells
> you which credential you have before you run. If Claude returns an org-policy
> error (HTTP 403), that's an account/organization limit, not a login problem — a
> different key or your org admin is the fix. And an outright `claude` CLI hang
> (e.g. a stale version) surfaces as a budget timeout, not an auth error.
A task is a plain Python function with **no body**; the signature and docstring
are the contract the agent fulfils at runtime — including its permissions:
`repo: sp.GitRepo` is the explicit writable workspace-handle grant that lets the
agent write the repository (see
[Permissions](#permissions-the-signature-is-the-permission-surface)):
```python
def write_program(
repo: sp.GitRepo,
prompt: str,Excerpt of 11,248 characters
Read on GitHub
100
Simon Yu · Northeastern University
11
6
2
1
Would you bet a product on this? Bounded 0–100 and slow moving.
matched fp:735bb9f61d652be7, topic:workflow-automation