Recently Evaluated

A macOS menu bar app that shows remaining Claude 5-hour and weekly usage limits locally, without reading tokens or calling an API.

A non-destructive compression gateway for AI coding agents that sits between the agent and the LLM API to reduce token usage while keeping original content recoverable on demand.

Bb is an agentic IDE that can control itself, with a desktop app, CLI, web app, and HTTP API for orchestrating coding agents and running work in live threads.

ARF defines a JSON record format for AI evaluation runs with canonicalization and reproducible SHA-256 digests, so scores can be traced back to the underlying question, model, inputs, and judgment.

XSAF is a small, fetch-native TypeScript framework for building secure AI agents. The page shows its core API, agent lifecycle, tool pipeline, MCP support, session handling, and offline testing story.

Preseason tracks which developer tools AI models recommend for a fixed panel of vibe-coding prompts, and publishes rankings across categories like auth, databases, queues, hosting, analytics, and LLM observability.

Browser UI for Claude Code, Codex, and Gemini CLI with generative UI, live cards, charts, pinned widgets, and local or BYO-key operation.

Open-source CLI that gives AI agents access to web reading, search, and platform-specific retrieval across sites like YouTube, Reddit, GitHub, Twitter/X, Bilibili, Xiaohongshu, and RSS.

Cloudflare’s Kitesurf is an agent-first browser built to run in V8 isolates on Workers, aimed at scalable browser execution for AI agents.

Open-source SDK for connecting AI agents to Slack, Microsoft Teams, Discord, and Telegram with native interactive UI and human approval flows.

A web calculator for estimating whether local LLMs fit on specific hardware and how fast they may run. It covers VRAM, token throughput, latency, power, and cost across multiple GPUs and devices.

A small macOS menu bar app that keeps your Mac awake so long-running AI agents can continue running with the lid closed.

A native macOS markdown writing app for plain .md notes on your own disk, with on-device AI search and semantic lookup.

A Hacker News front page clone that filters out AI-driven content and shows the remaining posts in a familiar HN layout.

Chrome extension that helps humans review AI-generated GitHub PRs by clustering diffs into review units and adding short AI summaries.

Open-source governance harness for AI agents with policy enforcement, human approvals, isolated execution, budgets, and signed audit evidence. It includes a CLI, REST API, web dashboard, SDKs, and adapters for common agent frameworks.

ModelPlane is an OpenAI-compatible AI gateway that routes requests across multiple model providers through one endpoint. It includes model groups, fallback and weighted routing, multiple provider support, reasoning output normalization, and usage/billing tracking.

A cross-platform CLI release for switching between isolated Claude Code accounts on Linux and macOS without re-login, with per-profile storage and Keychain-backed credentials on macOS.

RequestGuard creates CAPTCHA-gated short links that hide the destination until a visitor passes verification. It also supports expiry, passwords, and an API/SDK for generating protected links from a backend or embed flow.

An open-source project for running a small LLM and limited agent workflows on ESP32-P4 hardware, with model details, architecture notes, examples, and setup instructions.

BoundaryBench is an open-source benchmark for coding agents running under hardened sandbox policies. It compares harnesses like Claude Code, Codex, Terminus 2, and Grok on Terminal-Bench tasks and includes quickstart, policy controls, and result export tools.

Open-source skill contracts for AI agents that change response style based on the user’s state, such as overloaded, foggy, burnout, and reality-check modes. The README shows the intended behaviors, example prompts, and install instructions via Skills Store.

Lapse is a personal memory layer for AI tools, letting users speak naturally and have memories captured, tagged, and read back through MCP, a custom connector, and a REST API.

A read-only MCP server that lets AI assistants search, browse, and inspect public AI-agent listings from natural language inside clients like Claude and Cursor.