Topics / #agent-tools

Agent-tools projects on GitHub

Connectors and tool collections that give AI agents real-world capabilities.

As of , the most-starred Agent-tools projects on GitHub are composio (30.6K stars), xiaohongshu-mcp (16.3K stars) and lightcraft (13.2K stars). GitHub.Locker tracks 40 Agent-tools repositories with 135.1K stars combined. The fastest-growing on October 16 was xerj, with +1.2K stars in a day.

40Projects
135.1KStars in total
+2.2KStars in a day
Browse in the catalog →github.com/topics/agent-tools ↗

Hot right now

Most stars gained in a day · Oct 16

+1.2K

xerj

xerj-org/xerj · ★ 6.9K

XERJ is a local-first search engine that automatically indexes any folder - code, docs, logs, PDFs - so an AI coding agent can pull the exact passage it needs instead of grepping or reading whole files into context, and it speaks the Elasticsearch API so existing clients work against it unchanged.

+280

treg

superdesigndev/treg · ★ 5.7K

Treg is OpenRouter for agent tools: one base URL and one token let an agent call 3,000+ catalogued endpoints across 60+ providers — SEO, enrichment, ads, scraping, image and video generation — plus your team's own keys, skills and CLIs.

+262

monid

monid-ai/monid · ★ 4.7K

It's 'OpenRouter for agent tools': one base URL and one key give an agent access to 2,000+ tools across 72+ providers, so you integrate one API instead of wiring up dozens of vendors yourself.

+249

lightcraft

storytold/lightcraft · ★ 13.2K

LightCraft is an open-source, clean-room reimplementation of Adobe Lightroom built entirely in Rust: a non-destructive RAW photo editor and library manager with its own color pipeline, masking tools, and presets.

The all-time top

40 projects · by stars

  1. #1 · ComposioHQ/composio

    composio

    Composio is a platform for AI agents: it gives your agents 1000+ pre-authenticated toolkits, per-user sessions, auth, triggers, and a sandbox, so they can take action across apps instead of just chatting.

    30.6K
  2. #2 · xpzouying/xiaohongshu-mcp

    xiaohongshu-mcp

    An MCP server for Xiaohongshu (xiaohongshu.com) that lets AI assistants like Claude Code work with the platform directly: search and read posts, publish image or video content, comment, like and favorite.

    16.3K
  3. #3 · storytold/lightcraft

    lightcraft

    LightCraft is an open-source, clean-room reimplementation of Adobe Lightroom built entirely in Rust: a non-destructive RAW photo editor and library manager with its own color pipeline, masking tools, and presets.

    13.2K
  1. 4

    capsule-fs

    8.4Kunicity-aos/capsule-fs · Rust

    capsule-fs is a filesystem toolkit for AI agents inside Unicity AOS (also called Astrid OS): it gives agents readfile, writefile, replaceinfile, listdirectory, grepsearch, createdirectory, deletefile and movefile tools, all mediated by the OS kernel's VFS airlock.

  2. 5

    fast-jev-compaction

    7.7Ktamaratran/fast-jev-compaction · TypeScript

    A Claude Code plugin, also usable as an npm library, that replaces context compaction's summary with Jev decisions: every tool call and result is scored in fast requests, stale ones are dropped or truncated, and everything kept stays verbatim.

  3. 6

    xerj

    6.9Kxerj-org/xerj · Rust

    XERJ is a local-first search engine that automatically indexes any folder - code, docs, logs, PDFs - so an AI coding agent can pull the exact passage it needs instead of grepping or reading whole files into context, and it speaks the Elasticsearch API so existing clients work against it unchanged.

  4. 7

    semble

    6.2KMinishLab/semble · Python

    Semble is a code search library built for coding agents. It returns the exact relevant code snippets for a query instead of making the agent grep and read full files, using about 99% fewer tokens than grep+read.

  5. 8

    open-connector

    6Koomol-lab/open-connector · TypeScript

    OpenConnector is an open-source connector gateway that lets AI agents authenticate to and call 1,000+ SaaS providers (GitHub, Gmail, Notion, Slack, BigQuery, etc.) through a shared catalog of prebuilt Actions, acting as an open alternative to Pipedream or Composio.

  6. 9

    treg

    5.7Ksuperdesigndev/treg · Python

    Treg is OpenRouter for agent tools: one base URL and one token let an agent call 3,000+ catalogued endpoints across 60+ providers — SEO, enrichment, ads, scraping, image and video generation — plus your team's own keys, skills and CLIs.

  7. 10

    cli

    5.2Kentireio/cli · Go

    Entire is a CLI that hooks into your Git workflow to capture AI coding agent sessions (prompts, responses, files touched) and indexes them alongside your commits, creating a searchable record of how code was written.

  8. 11

    cli-printing-press

    4.8Kmvanhorn/cli-printing-press · Go

    CLI Printing Press is a generator that researches an API (its docs, every competing CLI, MCP server, and skill, or even a website with no public API) and then prints a token-efficient Go CLI, a matching MCP server, and a Claude Code skill for it, built specifically for AI agents to call thousands of times a day.

  9. 12

    monid

    4.7Kmonid-ai/monid · TypeScript

    It's 'OpenRouter for agent tools': one base URL and one key give an agent access to 2,000+ tools across 72+ providers, so you integrate one API instead of wiring up dozens of vendors yourself.

  10. 13

    just-bash

    4.4Kvercel-labs/just-bash · TypeScript

    just-bash is a simulated bash environment with a virtual filesystem, built so coding agents can run shell-like commands in a safe sandbox instead of touching the real file system. The repo describes the project as 'Bash for Agents'.

  11. 14

    mirage

    3.7Kstrukto-ai/mirage · TypeScript

    Mirage is a virtual terminal for AI agents that mounts services like S3, Google Drive, Slack, Gmail and Redis into one unified virtual filesystem, so agents access everything with familiar unix tools (ls, grep, find, jq), run virtual CLIs, route code to different runtimes, and stay inside a policy-controlled shell.

  12. 15

    agent-desktop

    1.8Klahfir/agent-desktop · Rust

    Agent Desktop is a Rust-built tool that gives AI agents reliable control over desktop apps by reading the OS accessibility tree instead of screenshots, so actions target real UI elements with stable, retry-safe references instead of guessed pixel coordinates.

  13. 16

    OpenLore

    1.3Kaakarim/OpenLore · Go

    OpenLore is a self-hosted, agent-native knowledge server: point it at a folder of Markdown and every agent on the team (Claude Code, Codex, Cursor, CI jobs) reads the same files over SSH and MCP, with control over who can see and change what.

  14. 17

    jev-browser-use

    1.2Kwy-coliney/jev-browser-use · JavaScript

    Jev Browser Use is a Skill for Codex (and soon Claude Code) that splits browser automation into two roles: a fast, cheap model called Jev handles clicks, navigation, toggles and scrolling, while Codex handles typing, visual judgment and the final verification. The split claims roughly 5-10x faster browser operations versus letting one model do everything.

  15. 18

    leviathan

    826elstongun/leviathan · Rust

    Leviathan turns your records (JSONL, JSON, CSV/TSV, SQLite, or anything a database CLI can export) into a single-file, ranked full-text search index so an AI agent gets the few records that answer a question instead of grepping through raw history.

  16. 19

    coop

    785trailofbits/coop · Rust

    coop is a Rust CLI that spins up disposable, isolated virtual machines so Claude Code and Codex can run with full tool access (Docker, git, compilers, package managers) without putting your host machine at risk.

  17. 20

    awesome-jev

    601kydlikebtc/awesome-jev · Python

    A curated index of 1207 public resources for Jev, TypeSafe AI's System One decision model, organised by the decision each resource illustrates. It's an independent catalog, not the product or an SDK.

  18. 21

    anysearch-dsh

    502anysearch-team/anysearch-dsh · TypeScript

    It's a plugin that connects the AnySearch search API to DeepSeek Harness (DSH), giving agents real-time web search, cleaned page content extraction, vertical/specialized search, and concurrent batch search.

  19. 22

    dsh-top100

    493evaldock-account/dsh-top100 · TypeScript

    dsh-top100 is a discovery, validation and trend index for the public plugin ecosystem of DeepSeek Harness (DSH): it ranks plugin repositories and catalogs Agent Skills by public GitHub signals.

  20. 23

    slashdocs

    380razorpay/slashdocs · Go

    slashdocs is a collaborative workspace where AI agents publish rich, self-contained HTML documents (reports, specs, dashboards, diagrams) and humans review them in the browser with anchored comments and versions.

  21. 24

    huashu-chrome

    339alchaincyf/huashu-chrome · JavaScript

    An MCP server + Chrome extension (22 browser tools) that lets AI agents like Claude Code, Codex CLI, Cursor and Gemini CLI drive your own Chrome with your existing logins — no API key, no re-login.

  22. 25

    DSH-taskboard

    336shengsheng90/DSH-taskboard · TypeScript

    DSH Taskboard is a plugin for DeepSeek Harness that adds native, local project and task management — SQLite-backed projects, tasks, comments, and workflows — plus tools that let AI agents claim and work on tasks under human review.

  23. 26

    dsh-ios

    326ZSeven-W/dsh-ios · TypeScript

    DSH iOS Simulator is a plugin for DSH (DeepSeek Harness) that puts a live iOS Simulator, or a USB-connected real iPhone, inside an AI agent conversation. It gives the agent 22 tools to boot devices, build and run Xcode projects, and drive the UI by accessibility identity, OCR text, or list-row taps, with the live screen streamed into an interactive sidebar panel.

  24. 27

    sentio

    254truespar/sentio · Rust

    Sentio is a complete multi-tenant email server for AI agents: it gives every agent its own real email address, delivers inbound mail as structured webhooks, and lets the agent reply in-thread over a REST API.

  25. 28

    agentplay

    223Entelligence-AI/agentplay · Python

    agentplay replays your coding agent sessions (Claude Code, Codex, Cursor, Copilot, OpenCode, Kiro) as an interactive map and timeline, showing what files it read or edited, what commands failed, and what it shipped.

  26. 29

    dsh-popout-sidebar

    208e2mcc/dsh-popout-sidebar · JavaScript

    A plugin for DeepSeek Harness (DSH) Web UI that adds a sidebar listing the agent's artifacts and a file tree, with multiple file preview types, and can pop out into a separate browser tab for viewing on another monitor.

  27. 30

    vomit

    200zachahn/vomit · Go

    Vomit converts Claude Code's token vomit into English by piping it through a local LLM. It hooks into Claude Code and rewrites its output right inside the Claude Code terminal UI.

  28. 31

    oto-dock

    191OtoDock/oto-dock · Python

    OtoDock is a self-hosted platform for a company's AI agents: your team builds them on Claude Code, Codex or local models, and they work in chats, on schedules, over the phone and on your own machines.

  29. 32

    dsh-android

    185ZSeven-W/dsh-android · TypeScript

    DSH Android is a plugin for DeepSeek Harness (DSH) that puts a live Android device — an emulator or a USB-connected phone — inside an AI agent conversation, letting the agent view, control, and inspect it entirely through adb.

  30. 33

    lazy-frames

    165cosmicstack-labs/lazy-frames · TypeScript

    Lazy Frames is an agentic video generation tool: you feed a skill URL to your AI coding agent, which captures source material, authors a spec, and renders it into a pixel-perfect, byte-stable MP4, with the core pipeline running fully locally and no cloud required.

  31. 34

    technocore-chat

    165flop-labs/technocore-chat · Python

    technocore-chat is a zero-auth chat and notes service built specifically for AI agents whose sandboxes only allow fetching URLs: every operation, including writes, is a single plain GET request. It runs live at technocore.chat and can also be self-hosted.

  32. 35

    toolrush

    157OnlyTerp/toolrush · Python

    ToolRush is a low-overhead execution layer plugin for Hermes Agent (Nous Research's agent harness) that eliminates the latency overhead of tool calls like file reads, searches, and terminal commands while keeping the same tools, outputs, and safety gates.

  33. 36

    Matryoshka

    150yogthos/Matryoshka · TypeScript

    Matryoshka (RLM) is an MCP server and CLI toolkit that lets LLM agents analyze documents up to 100x larger than their context window, by having the model emit symbolic 'Nucleus' commands that a logic engine called Lattice executes server-side, instead of chunking text into a vector database.

  34. 37

    orchflows

    121DanMcInerney/orchflows · Python

    Orchflows gives you two composable skills for AI coding agents that let you build a reusable multi-step workflow once and reuse it inside bigger jobs instead of re-explaining the process every session.

  35. 38

    calnode

    118Calnode/calnode · Go

    Calnode is a self-hostable scheduling app like Calendly or Cal.com, shipped as a single Go binary with an embedded SQLite database. It includes first-party video meetings, recording, AI notetaking, and is built API-first with a native MCP server for AI agents.

  36. 39

    voiceshell-muse-bridge

    77hyj-STAR/voiceshell-muse-bridge · Python

    VoiceShell Muse Bridge is an experimental research project that lets an ESP32 board pair with Meta's Muse Gadget SDK, then hands the actual text conversation off to a separate Linux server instead of routing every message through the board.

  37. 40

    dsh-antibrow

    11antibrow/dsh-antibrow · JavaScript

    dsh-antibrow is a DeepSeek Harness plugin that gives an AI agent a persistent browser identity through AntiBrow, a Chromium fork with engine-level fingerprint spoofing. It exposes 16 MCP tools covering sessions, profiles, proxies and page control, driven via the standard Playwright API, with optional residential proxy egress.

Often tagged together