Trending

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.

Hotpath records an AI agent run once, compiles the trace into an editable workflow, and replays it with only the judgment-heavy steps still using an LLM.

Open-weight System One model for machine-intuition style judgment tasks, built on a T5Gemma 2 encoder-decoder base. The page shows what the model does, how it is used in Python and via CLI/HTTP API, and includes benchmark and example sections.

Shopify documentation for browser-based AI agents to read, update, and complete checkout flows in the buyer’s browser using WebMCP.

Read-only Postgres MCP server that adds schema context files for more accurate agent answers and supports agents writing back table notes.

A working example of an AI helpdesk agent built on the Phaze Enterprise API and Phaze MCP. It watches Slack requests, connects to Windows machines, and can hand off to a human when needed.

A local-first macOS desktop app for tracking manga and book libraries and turning chapters into narrated recap videos using local vision models, TTS, and swappable video backends.

Jylus is an API for turning changing records, logs, telemetry, and JSON into compact, source-backed evidence packs for AI systems. The page includes a free trial, API examples, sample response structure, and published benchmark results.

RTSP stream simulator for developers and AI teams to test computer vision, video analytics, and VMS integrations using MP4 files instead of physical cameras.

Open-source staff scheduling platform with a native MCP server so AI agents can search candidates, create assignments, manage shifts, and handle staffing requests.

A llama.cpp fork that adds a JEV-compatible decision API for Choice, Score, and Noul evaluations while keeping standard LLM inference support.

Open-source, self-hosted platform for running managed AI agents like Claude Code, Codex, Hermes, DeepSeek Harness, and Pi on infrastructure you control.

Open-source demo for recognizing hand-object manipulation stages in everyday video, with installation steps, run commands, and a short explanation of the pipeline.

OpenAPPA is an open-source deterministic guardrail and information-flow policy engine for LLM agents. The page describes flow-based security labeling, CI validation, remedy plans, sanitizers, authorities, and subagents, with links to docs, GitHub, Discord, and benchmark claims.

A Vercel article showing how to use GPT-6 Sol with the AI SDK evaluation API to assess draft replies, score evidence-based answers, and read structured results in TypeScript.

AI code security scanner that reviews a repository, flags exploitable issues, and provides a scoped fix prompt for each finding.

An offline hybrid retrieval engine for RAG that runs on-device, including phones. The repo shows BM25 plus dense retrieval, cross-encoder reranking, sparse expansion support, and bindings for Rust, Python, Swift, and Kotlin.

Focal Prompt is an analysis lab for studying how prompts and prompt sections influence model behavior. It includes focus labeling, prospective and retrospective focus prediction, baseline output sampling, ablation tests, batch analysis, and an agent builder.

An in-memory Tree-sitter AST pruning and context firewall MCP server for coding agents like Cursor and Claude. The repo includes installation steps, diagnostics, benchmarks, and a clear usage path for local setup.

A LangGraph-based workflow that generates, benchmarks, validates, and iterates CUDA kernels using a C++/CUDA harness. It targets workload-specific optimization with support for NVIDIA research lookup and optional Nsight Compute profiling.

A local gateway for Claude Code that logs tool calls, enforces approval rules, and trims noisy tool output to save tokens.

A Mac notch app that surfaces Claude Code and Codex approval prompts in the menu bar/notch so developers can approve or deny requests without switching back to the terminal.

A GitHub repo benchmarking TypeSafe’s Jev against Sonnet, GPT-5 nano, and local models on 770 Reddit AITA verdicts, with Brier scores, latency, and cost comparisons.