Recently Evaluated

Python library for managing LLM conversation history with cache-friendly message ordering, dynamic context, and compaction.

A 150M ModernBERT-based token classification model that extracts verbatim evidence spans for RAG and related pipelines, reducing the need for an LLM call to highlight answers.

Foyer is a local dashboard for watching AI coding agents work, with a narrated current-focus view and a research panel for reading sourced briefings while you wait.

Workplane lets users publish Markdown, HTML, PDFs, and screenshots as shareable pages for humans and agents, with comments, versioning, and MCP-based publishing from AI tools.

Open-source project that ports Google DeepMind’s Magenta RealTime 2 music generation to Core ML for local execution on iPhone, with Apple Neural Engine acceleration and no GPU use.

HelixDB is an OLTP graph-vector database built in Rust, aimed at AI apps, knowledge graphs, and memory layers. The page shows the product’s core model, CLI workflow, SDK examples, and local setup commands.

Blue41 is an enterprise risk control platform for AI agents. The site says it monitors agent behavior in production, detects prompt-injection-style incidents and unauthorized activity, and helps teams control sensitive workflows and compliance risk.

Nodea is a live branching AI chat canvas for working through complex projects. The demo shows forkable replies that turn conversations into trees, with example prompts for naming a coffee shop and shaping branding ideas.

Open-source AI-powered desktop automation agent that gives an LLM a fully isolated Ubuntu workspace with browser-based supervision, MCP access, audit logs, and OS-level control.

Chrome extension that tracks major AI companies, pulls in AI news and research, and generates daily summaries with Gemini, including language-aware summaries and article deep dives.

Open-source relay for user-authorized AI sessions. Lets app builders connect users to ChatGPT, Grok, or GitHub Copilot subscriptions via OAuth, then call models through a relay using the OpenAI SDK or a React component.

RiskKernel is an open-source, self-hosted reliability runtime for AI agents. It adds hard budgets, kill switches, crash-resumable runs, human approval gates, and OpenTelemetry export, with support for OpenAI, Anthropic, and existing agent stacks.

A free, Apple-notarized macOS menu-bar app that forces short rain-themed breaks while AI agents keep running in the background.

LocalCode turns plain English into CLI commands on Apple Silicon Macs using Apple’s on-device Foundation Models framework, then asks for approval before running them.

SlopGuard is a desktop Chrome extension that scans images on webpages for AI generation signals using local detection methods, including metadata, SynthID, and a visual classifier.

A Steel blog post that dissects Claude Code’s deep-research workflow and contrasts it with other research systems. It also positions Steel as an open-source browser API for AI agents, making the page relevant for AI builders.

AutoGPU is an open-source project for using AI agents to design and harden a GPU-like chip end to end, from Verilog through synthesis and place-and-route to GDSII. The repo shows a concrete fp8 matmul accelerator flow, quick-start steps, and documentation for running the design/hardening process.

Linux-only kernel-level audit feed for Claude Code or other matched agent sessions, streaming exec, file, and network activity in real time.

Runtime safety net for LLM agents that detects runaway token growth, classifies failure patterns, and suggests fixes without extra LLM calls. The project includes a Rust core, Python SDK, examples, benchmarks, and benchmark claims across multiple models and runs.

Open-source voice-driven multi-agent assistant for building real projects on a Mac, with voice, web chat, and WhatsApp channels plus CLI-based task runners.

A Blue41 case study on how a banking AI assistant could be abused through indirect prompt injection, and what mitigation layers help reduce the risk.

A macOS SwiftBar plugin that shows Claude Code quota gauges in the menu bar, with per-account usage, reset timers, and a dropdown with detailed limits.

A set of small AI app experiments focused on “the use of the useless,” including AI Almanac, Forget, and Nickname Machine for history, closure, and naming moments.

A focused article-style site arguing for domain-specific languages in LLM workflows, with concrete examples around structured extraction, validation, and pricing rules.