
Sign up to save your podcasts
Or


Alex and Jordan break down 6 new AI repos climbing GitHub today: firecrawl, knowledge infrastructure for AI agents; opensre, an agent that investigates failed CI runs; app-store-compliance, a pre-submission guard against App Store rejections; autoclip, an AI video clipper that turns long podcasts into short clips; ffmpeg-skill, a local video editor for coding agents; and instagram-agent-skill, thirteen Claude skills that run an Instagram account.
Repos:
firecrawl/firecrawl — https://github.com/firecrawl/firecrawl
Tracer-Cloud/opensre — https://github.com/Tracer-Cloud/opensre
mjmirza/app-store-compliance — https://github.com/mjmirza/app-store-compliance
zhouxiaoka/autoclip — https://github.com/zhouxiaoka/autoclip
kajisho5/ffmpeg-skill — https://github.com/kajisho5/ffmpeg-skill
Jakeschincariol/instagram-agent-skill — https://github.com/Jakeschincariol/instagram-agent-skill
Alex and Jordan break down 3 new AI repos climbing GitHub today: n8n-io/n8n, a fair-code workflow automation platform with native AI capabilities for building AI agents and multi-step workflows on a visual canvas; punkpeye/awesome-mcp-servers, a curated collection of MCP servers organized by category so AI assistants can plug into databases, browsers, and coding tools; and openvinotoolkit/openvino, Intel's open-source toolkit for optimizing and deploying AI inference so models run faster on ordinary CPUs, GPUs, and AI accelerators.
Repos:
n8n-io/n8n — https://github.com/n8n-io/n8n
punkpeye/awesome-mcp-servers — https://github.com/punkpeye/awesome-mcp-servers
openvinotoolkit/openvino — https://github.com/openvinotoolkit/openvino
Alex and Jordan break down 4 new AI repos climbing GitHub today: mvanhorn/last30days-skill, an AI agent skill that researches any topic across Reddit, X, YouTube, Hacker News, Polymarket, and the web, then synthesizes a grounded summary; backnotprop/plannotator, a local browser workspace for staying in the loop on your coding agents; jingyaogong/minimind, a way to train a small language model from scratch in about two hours on a single graphics card; and google-research/paper-orchestra, a multi-agent framework from Google Research that turns raw research materials into a submission-ready paper.
Repos:
mvanhorn/last30days-skill — https://github.com/mvanhorn/last30days-skill
backnotprop/plannotator — https://github.com/backnotprop/plannotator
jingyaogong/minimind — https://github.com/jingyaogong/minimind
google-research/paper-orchestra — https://github.com/google-research/paper-orchestra
Alex and Jordan break down 4 new AI repos climbing GitHub today: google-labs-code/stitch-skills, a library of Agent Skills that teaches coding agents to design and build app screens; ggml-org/ggml, the plain-C tensor library behind llama.cpp for running machine learning models locally; elementalsouls/Claude-BugHunter, a Claude Code skill bundle for authorized bug-hunting built from 681 real disclosed reports; and NVIDIA/SkillEvaluator, a three-tier framework for grading AI agent skills for safety, overlap, and live effectiveness.
Repos:
google-labs-code/stitch-skills — https://github.com/google-labs-code/stitch-skills
ggml-org/ggml — https://github.com/ggml-org/ggml
elementalsouls/Claude-BugHunter — https://github.com/elementalsouls/Claude-BugHunter
NVIDIA/SkillEvaluator — https://github.com/NVIDIA/SkillEvaluator
Alex and Jordan break down 3 new AI repos climbing GitHub today: ag-ui-protocol/ag-ui, an open protocol that plugs AI agents into the apps people already use; Tencent/LoopForge, a self-hosted workspace where a team ships software with coding agents; and dream-num/univer, an open-source Office SDK for embedding spreadsheets, docs, and presentations into your own product.
Repos:
ag-ui-protocol/ag-ui — https://github.com/ag-ui-protocol/ag-ui
Tencent/LoopForge — https://github.com/Tencent/LoopForge
dream-num/univer — https://github.com/dream-num/univer
Alex and Jordan break down 2 new AI repos climbing GitHub today: google/elephant-goldfish-model, Google's open-source plugin for designing and reviewing software with AI agents; and Vincentwei1021/video-shotcraft, a skill that turns a product into a polished promo film.
Repos:
google/elephant-goldfish-model — https://github.com/google/elephant-goldfish-model
Vincentwei1021/video-shotcraft — https://github.com/Vincentwei1021/video-shotcraft
Alex and Jordan break down 5 new AI repos climbing GitHub today: GoogleChrome/modern-web-guidance-src, skills that keep your coding agent up to date with the latest web best practices; backnotprop/plannotator, a local workspace to stay in the loop with your coding agents; tradermonty/claude-trading-skills, Claude Code skills for investors and traders; ChromeDevTools/chrome-devtools-mcp, Chrome DevTools for coding agents; and lobehub/lobehub, super-agents that join your team's chats.
Repos:
GoogleChrome/modern-web-guidance-src — https://github.com/GoogleChrome/modern-web-guidance-src
backnotprop/plannotator — https://github.com/backnotprop/plannotator
tradermonty/claude-trading-skills — https://github.com/tradermonty/claude-trading-skills
ChromeDevTools/chrome-devtools-mcp — https://github.com/ChromeDevTools/chrome-devtools-mcp
lobehub/lobehub — https://github.com/lobehub/lobehub
Alex and Jordan break down four new AI repos climbing GitHub today: Qwen Code, Alibaba's open-source terminal coding agent that works with almost any model; TensorFold, a tiny Zig binary that runs language models on your Mac with no Python setup; Free Claude Code, a switchboard routing a dozen coding assistants to 61 model providers including free tiers; and llama.cpp, the foundational engine for running language models locally on almost any hardware.
Repos:
QwenLM/qwen-code — https://github.com/QwenLM/qwen-code
ashhart/TensorFold — https://github.com/ashhart/TensorFold
Alishahryar1/free-claude-code — https://github.com/Alishahryar1/free-claude-code
ggml-org/llama.cpp — https://github.com/ggml-org/llama.cpp
Alex and Jordan break down three new AI repos climbing GitHub today: improve, an agent skill from shadcn that audits your codebase with your smartest model and writes step-by-step plans for cheaper models to run; codex-security, OpenAI's scanner that finds, validates, and fixes vulnerabilities in your code; and herdr, a terminal workspace that keeps your coding agent sessions running, flags the stuck ones, and works across machines.
Repos:
shadcn/improve — https://github.com/shadcn/improve
openai/codex-security — https://github.com/openai/codex-security
herdrdev/herdr — https://github.com/herdrdev/herdr
Alex and Jordan break down four new AI repos climbing GitHub today: poki, which lets an AI agent operate a real iPhone through a local MCP server; devvm, one isolated VM per project for running coding agents in a sandbox; uncut, a tiny headless coding-agent harness in a single 7 MB binary; and nvke, NVIDIA's management plane for running managed Kubernetes clusters on your own GPU servers.
Repos:
JosssphZhou/poki — https://github.com/JosssphZhou/poki
devvmhq/devvm — https://github.com/devvmhq/devvm
Balanov/uncut — https://github.com/Balanov/uncut
dsx-ai-factory/nvke — https://github.com/dsx-ai-factory/nvke