Trending

A web tool for estimating how fast local LLMs will run on different hardware before you buy it. It lets users browse by GPU, Mac, CPU, and model, and shows simulated tokens/sec and time-to-first-token based on benchmark data.

A native Android app that generates full songs with vocals on-device using YuE2, with local storage, offline use, stem splitting, a mixer, recording tools, and AI writing aids.

A guardrails and knowledge-layer tool for coding agents. It briefs agents before tasks, verifies rules with checks, and flags drift when governed files change. Works with Claude Code and other MCP agents.

A Claude Code mod that adds debugger-like controls for tool calls, including breakpoints, step/reject, timeline inspection, and failure diagnosis.

A macOS menu bar app for monitoring and approving AI coding agents like Claude Code, Codex, Gemini CLI, OpenCode, OpenClaw, and Hermes. It shows live status, last messages, quick replies, and one-click approval, and it can build from source as a free open-source project.

Rinkata is a shared project context layer for teams using coding agents, with knowledge capture, goal tracking, decision handling, and evidence packs for completed work.

Botscent is a lightweight library that detects AI agents visiting a site and can identify which agent is behind a request when evidence is available. It includes both a server-side signal check and a small client-side script, with docs, FAQ, and GitHub links shown on the page.

Favz indexes public GitHub repos to show which MCP servers, skills, plugins, and agent configs people actually set up. It tracks rising tools, compares stars with real usage, and includes a CLI/plugin angle for monitoring your own setup.

Open-source semantic layer for warehouses that infers table, column, join, and business-rule metadata, then serves it to AI agents through MCP and CLI checks.

A reverse proxy for OpenAI-compatible agent traffic that detects runaway tool loops and enforces hourly and daily spend caps with zero payload retention.

Open-source personal agent for phone, desktop, and browser that can take actions, keep running in the background, remember user preferences, and coordinate across devices.

OpenChart is an open-source trading and charting app that adds AI agents to market analysis, alerts, and indicator workflows. It supports custom logic in Tea, source-linked chart annotations, and agent-assisted research using models like Claude, Codex, and Gemini.

A GitHub repository for starting, watching, and stopping a Claude Code session on a Windows PC from a phone over Tailscale. The README explains the architecture, setup steps, requirements, and security tradeoffs in detail.

An open-source browser extension for creating and managing userscripts by chatting with an LLM. It inspects the current page, writes and tests scripts live, supports multiple model providers, and can export, import, and run mods across supported browsers.

A detailed technical post about IronBee Express and its browser-agent architecture, showing how Jev is used for fast, low-cost step selection and when an LLM takes over.

Project-local MCP server that lets existing agents discover each other, share status, and exchange durable direct messages with explicit acknowledgement.

Recly turns an Apple Watch or Galaxy Watch into a recorder, transcribes locally or with your own API key, and stores files in your Google Drive with no Recly server in the middle.

A tiny programming language for AI agents that bakes in permissions, bounded retries, verification, and a tamper-evident audit log.

An MCP server that converts public URLs into clean Markdown for AI agents and MCP-compatible clients. The repo includes setup instructions, a usage example, a direct API option, and notes on supported clients like Claude Desktop and Cursor.

A repo for keeping coding agents from overengineering by injecting a ruleset during coding and running a slop-check after edits. It supports multiple agent setups, including Claude Code, Codex, OpenCode, Cursor, Copilot, and generic AGENTS.md-based workflows.

A command-line tool that sends fake RFC 3464 bounce messages to make spam senders think an address does not exist. The README explains the workflow, configuration, batch mode, multiple-account handling, and provider-specific caveats.

Kodama is a portable, provider-neutral multi-agent pack for the Kiro CLI. It ships one orchestrator plus eight specialist agents, with install options for Kiro Crew or standalone use, and includes CLI commands for memory, health checks, updates, and uninstall.

Open-source terminal app for multi-agent coding that lets one lead agent coordinate a team, manage worktrees, and keep the user in one thread.

SundaeDB is an agent-first analytics tool that lets AI agents query CSV and ZIP data in plain language, with budget caps, pay-per-query pricing, and partial results when limits are reached.