Show HN
amanharshx /
ultralytics-mcp
Control YOLO Training and Datasets from Claude/Cursor via MCP
好代码，值得被发现。
下一个
宝藏项目.
开源项目、精巧工具，还有让人眼前一亮的点子。
从 Hacker News 社区讨论中，找到你的下一份灵感。
Y社区的眼光，你的好奇心。无付费推荐。
2,872独立仓库
4,114Show HN 帖子¹
3,330GitHub 外链¹
2026 年 7 月快照
社区发现的互联网一角
创造，然后分享。✳
按 HN 分数排序，非 GitHub Star 数
Show HN
mihirahuja1 /
agentnestOSS
AgentNest, self-hosted sandboxes for AI agents
Show HN
projectargus-cc /
libargus.cc
Low-latency local LLM runner via OpenJDK Panama FFM (Java 22)
Show HN
mazen160 /
backlog
Backlog – tasks and contexts manager for AI coding agents
Show HN
samvallad33 /
vestige
Vestige – A queryless memory layer for multi agent coding fleets
Show HN
theoriclabs /
agent-convert
You can continue a Claude session in Codex, and vice versa
Show HN
itdar /
newsline
Newsline – one-line news in your status line while your AI agent works
Show HN
PJHkorea /
Continuous_Wave_Field_LLM_Brain
Autograd-Free LLM Guiding with 0MB VRAM (Alternative Pathways)
Show HN
cyrusNuevoDia /
gpt-workflow
Claude Code Workflows for GPT/Codex
Show HN
Rudekwydra /
burnless
Burnless – 1,590 tokens of context for a 1.44M-token workday
Show HN
lucasgodt /
open-agent-ready-typescript
A TypeScript repo where AI agents can't break the architecture
Show HN
mohit17mor /
SquadAI
SquadAI is the Kubernetes-like control plane for Codex agents
Show HN
Mojo0869 /
ABSL
ABSL – 100% Integer-Only Neural Network Training in Rust
Show HN
ecrasevvv /
c-fm
C-Flow Matching Inference
Show HN
agileAlligator /
mcploitable
Mcploitable – Vulnerable MCP Servers for the OWASP Agentic Top
Show HN
overflowy /
herdr-claude-gpt-adversarial-review-skill
Adversarial code review setup with herdr, Claude and GPT-5.6-sol
Show HN
PJHkorea /
pim-hbm-bypass
HW/SW co-design blueprint for fault recovery in LLM training
Show HN
tarunlnmiit /
autopilot-jobhunt
Agent scans 130 careers pages nightly, scores jobs vs. your resume