GitHub trending, June 8, 2026: agent 'skills' lead the board

Fastest-growing this period

AI agent skill that researches any topic across Reddit, X, YouTube, HN, Polymarket, and the web - then synthesizes a grounded summary

+141

Leonxlnx/taste-skill Shell

Taste-Skill - gives your AI good taste. stops the AI from generating boring, generic slop

+80

pewdiepie-archdaemon/odysseus Python

Self-hosted AI workspace.

+73

chopratejas/headroom Python

Compress tool outputs, logs, files, and RAG chunks before they reach the LLM. 60-95% fewer tokens, same answers. Library, proxy, MCP server.

+71

RyanCodrai/turbovec Python

A vector index built on TurboQuant, written in Rust with Python bindings

+66

NousResearch/hermes-agent Python

The agent that grows with you

+57

safishamsi/graphify Python

AI coding assistant skill (Claude Code, Codex, OpenCode, Cursor, Gemini CLI, and more). Turn any folder of code, SQL schemas, R scripts, shell scripts, docs, papers, images, or videos into a queryable knowledge graph. App code + database schema + infrastructure in one graph.

+55

colbymchenry/codegraph TypeScript

Pre-indexed code knowledge graph for Claude Code, Codex, Gemini, Cursor, OpenCode, AntiGravity, Kiro, and Hermes Agent — fewer tokens, fewer tool calls, 100% local

+45

farion1231/cc-switch Rust

A cross-platform desktop All-in-One assistant for Claude Code, Codex, OpenCode, OpenClaw, Gemini CLI & Hermes Agent. Only official website: ccswitch.io

+44

microsoft/markitdown Python

Python tool for converting files and office documents to Markdown.

+44

Our read

Today’s board is led by mvanhorn/last30days-skill, an agent skill that researches a topic across Reddit, X, YouTube, HN, and the web before synthesizing a grounded summary — the second “skill” repository in the day’s top two, alongside Leonxlnx/taste-skill. The packaging of agent behavior as a distributable skill is clearly having a moment.

The outlier is RyanCodrai/turbovec at number five: a vector index built on TurboQuant, written in Rust with Python bindings. It is the day’s most prominent piece of pure infrastructure rather than agent tooling — a reminder that the performance layer underneath all this AI plumbing is still drawing its own builders.