Recently Evaluated

Hyperset is a local-first debugging tool for coding agents that captures runtime evidence from your app and feeds it back into the agent. It supports Node.js, Python, and Go, and can work with its own agent or tools like Claude Code, Codex, and OpenCode.

Open-source MCP proxy that blocks agent send/share actions when the destination first appears in untrusted tool output. It uses deterministic rules rather than an LLM and logs every allow/deny decision.

A client for a neutral AI referee service that handles tie-breaks, sealed reviews, and signed receipts for AI agents, with pay-per-request billing over x402 on Base.

A defensive AI auditing API for checking structural coherence, semantic rupture, and interaction stability, with a public demo, API docs, SDK install steps, and paid plans.

A research write-up about a system that tracks Reddit stock chatter, replays LLM buy/hold/sell calls, and compares them with later price moves and simple baselines.

A repository that decodes braille music notation into a performance of Gottschalk’s Grande Tarantelle, with Python scripts to parse the score, generate MIDI, and render audio.

A Rust-based MCP server that lets AI agents chat with each other, manage channels, and be monitored through a web page.

A data-backed study showing which major websites block specific AI crawlers in robots.txt, with breakdowns by crawler, site type, and a downloadable CSV.

Smelt is an open-source terminal coding agent with a Rust core and Lua plugins, aimed at developers who want a fast, customizable command-line workflow. The page shows install steps, plugin support, vim mode, themes, and benchmark claims comparing memory and speed against other agents.

A Java library for adding privacy and security guardrails to Spring AI apps, including PII tokenization, tool authorization, and content inspection.

Bifrost’s Agent Gateway adds A2A agent calls to its existing gateway, with logs, access control, transport bridging, and callback handling across models, tools, and agents.

A spec-driven workflow for AI coding agents that uses reviewed specs, isolated worktrees, and gated commits to turn ideas into verified pull requests.

A technical write-up on packing hundreds of AI agent sandboxes onto limited hardware, covering VM booting, idle killing, snapshots, and routing lighter tasks outside the VM.

An agent skill for coding assistants that helps them produce more deliberate page design instead of generic layouts. It includes install commands, example outputs, supported tools, and a concrete workflow for planning, reviewing, and refining UI.

Self-hosted AI code reviewer for GitHub pull requests that uses multiple narrow reviewer passes and a validator to produce one verified review.

Hob is an open-source desktop AI assistant that runs locally on Mac, Windows, and Linux. It positions itself as a private, offline alternative to ChatGPT and Claude, with support for open-weight models like Gemma and gpt-oss.

A crowdsourced benchmark that asks users to upload photos without obvious toilet signs and label where they would go to find the restroom, aiming to test whether AI can handle this kind of real-world visual inference.

Claix is a hosted controlled database retrieval layer for AI agents. It lets agents query Postgres or Supabase through read-only, route-based access without writing SQL, and exposes endpoints over REST, MCP, and A2A.

A data-driven Equibles research piece analyzing 5,305 S&P 500 earnings calls to track how often companies say “agentic,” “generative AI,” and AI developer names like OpenAI and Anthropic. The page includes a summary, methods, and supporting tables/citation info.

DocsGPT documents a small classifier used for bounded decisions like prompt-injection checks, relevance scoring, and conversation analytics. It explains where the model runs, how it is enabled, what endpoints it uses, and how it behaves on CPU, Mac GPU, or NVIDIA GPU.

Tomte is a Go toolkit for building small, single-purpose AI agents that run on a schedule or on demand. The page shows deployment, runtime limits, example code, and a concrete invoice-chasing use case.

Drint provides an MCP server for packaging and print design workflows, letting assistants like Claude, ChatGPT, and Muse search mockups, open 3D views, and generate exact-size dielines.

A CLI tool that gives coding agents names from a roster so teams can track who is working on what across Claude Code, Codex, tmux, and cloud sessions.

An MCP server and CLI for AI agents to write, schedule, publish, and review Threads posts through Meta’s Threads API, with permission controls and support for Claude, Cursor, and VS Code.