GitHub Trending Weekly

GitHub Repo Briefing

Twelve weekly trending repositories, summarized by what they do and how to test them
May 23, 2026
Issue 05
01
Agent Ops

codegraph

colbymchenry/codegraph

A local code knowledge graph that indexes a repository, syncs after code changes, and serves context to coding agents through CLI and MCP clients.

18,340
Stars total / TypeScript
14,072 stars this week

A first test

Run codegraph init in a disposable clone and index one module. Ask for a caller list, compare it with the language tools, and check what files the agent can read before enabling edits.

02
Agent Ops

openhuman

tinyhumansai/openhuman

A local-first personal AI system that combines memory, agent orchestration, workflows, and deep research in one desktop application.

26,105
Stars total / Rust
17,124 stars this week

A first test

Install it without personal data and run one research task using public sources. Inspect the memory store and outbound requests before importing notes or connecting accounts.

03
Agent Ops

academic-research-skills

Imbad0202/academic-research-skills

A Claude Code skill workflow for academic research: plan the question, gather sources, write, review, revise, and finalize with a human in the loop.

19,484
Stars total / Python
11,550 stars this week

A first test

Install it and run /ars-plan for one narrow question. Check the source list and claims after each stage; do not accept a final draft that lacks evidence for a key statement.

04
Creative Systems

RuView

ruvnet/RuView

A WiFi-sensing system that estimates presence, movement, and some vital-sign signals from radio data without using a camera.

64,482
Stars total / Rust
6,773 stars this week

A first test

Run ruview doctor and ruview verify in a test room. Compare the output with a known movement test and treat any health-related signal as experimental, not a medical measurement.

05
Agent Ops

agentmemory

rohitg00/agentmemory

A persistent memory service for coding agents and MCP clients. It stores project context and exposes a demo server for checking how memory is added and retrieved.

16,659
Stars total / TypeScript
6,891 stars this week

A first test

Run the setup and agentmemory demo --serve in a disposable project. Store one fact, retrieve it in a new session, and verify the record before adding private notes.

06
Open Source

supertonic

supertone-inc/supertonic

An ONNX text-to-speech engine that runs on-device and supports multilingual speech generation across 31 languages.

9,670
Stars total / Swift
3,621 stars this week

A first test

Install it with pip install supertonic and synthesize one short sentence in two languages. Check latency, voice quality, output format, and whether the model runs without a network call.

07
Open Source

CloakBrowser

CloakHQ/CloakBrowser

A source-patched Chromium build intended as a drop-in Playwright or Puppeteer browser for controlled browser testing and automation.

19,289
Stars total / Python
7,042 stars this week

A first test

Run the supplied browser example against a local test page. Review the security, compliance, and site terms first; do not use it to bypass access controls or anti-abuse systems.

08
Creative Systems

ViMax

HKUDS/ViMax

An agentic video-generation workflow that separates director, screenwriter, producer, and video-generator roles before producing a video.

6,856
Stars total / Python
2,685 stars this week

A first test

Run one short prompt with test media and save each stage's output. Check the script, asset sources, and final edit before using the video in a public channel.

09
Agent Ops

12-factor-agents

humanlayer/12-factor-agents

A set of engineering principles and examples for building reliable LLM applications, including explicit context, control flow, tools, and human approval.

21,843
Stars total / TypeScript
1,907 stars this week

A first test

Choose one principle and apply it to a small agent. Add a test that proves the agent pauses or fails at the required boundary instead of relying on a prompt alone.

10
Open Source

ai-engineering-from-scratch

rohitg00/ai-engineering-from-scratch

A hands-on course that teaches AI engineering by building and shipping small projects with an agent as a tutor.

13,032
Stars total / Python
3,715 stars this week

A first test

Install its skills and run /start-learning for one topic. Make the smallest exercise yourself, run its test, and record what you changed before asking the agent for the next lesson.

11
Open Source

easy-vibe

datawhalechina/easy-vibe

A step-by-step course for beginners who want to build modern software with AI-assisted development and ship working projects.

14,164
Stars total / JavaScript
2,942 stars this week

A first test

Start one lesson in a clean project and complete its exercise without skipping the verification step. Keep the generated code only after you can explain its entry point and test.

12
Agent Ops

Understand-Anything

Egonex-AI/Understand-Anything

A coding-agent plugin that analyzes a repository into an interactive knowledge graph, dashboard, and searchable explanations.

20,188
Stars total / TypeScript
4,680 stars this week

A first test

Install it in a disposable clone and run /understand. Compare one graph relationship with the source code, then use /understand-diff before accepting a change.

End of Issue
Check the links. Test one repo.
GitHub Trending Weekly / Issue 05
Generated by Codex from SnailDev github-hot-hub weekly archive: https://raw.githubusercontent.com/SnailDev/github-hot-hub/main/archives/2026-05-23.md