Search

TypeScript SDK for Torrix, self-hosted observability for AI agents: record OpenAI, Anthropic and Vercel AI SDK calls with zero dependencies

Lightweight browser automation CLI for AI agents, powered by Lightpanda (fork of agent-browser)

Universal, local-first, persistent memory layer for AI agents, LLM apps, and chatbots.

Time-travel debugging and causal RCA for LangGraph agent traces, on a causal graph.

TypeScript AI agent runtime. Write your agents in code and choose how predictable they are: deterministic steps where they can be, an LLM only where a step needs judgement, all on your machine.

Distill any LLM into a small, fast, calibrated classifier. A teacher LLM (Claude, Gemini, OpenAI or any compatible API) labels your data with probabilities, a multilingual encoder learns them, and a FastAPI service answers yes/no, choice and score questions in milliseconds, no LLM calls at inference. By Salto Ángel.

Local observability for Claude Code: sessions, subagents, blocks, logs, tokens and cost on OpenTelemetry, Prometheus, Loki and Grafana.

Karpathy-pattern LLM wiki maintainer: ingest, promote, query, lint and audit a curated markdown knowledge base.

Zero-auth deploy primitive. Turns any folder into a live URL in one command. For AI agents and humans.

三生万物 —— 极简 LLM Agent Harness，agent 天然自我演化

Autonomous AI Search, Training & Agent Crawler Diagnostic Engine

Typed synchronous and asynchronous Python access to the AgentHealth CLI

Secure prompt engineering for AI agents and engineering teams

This model always redirects to the latest model in the Pareto family.

Pareto is a multimodal composite model built for research, coding, and agentic workflows, while delivering frontier-level performance across a broad range of general-purpose tasks.

Inventory and clean up what coding-agent CLIs (Codex, Claude Code, Cursor, Gemini, ...) leave behind on macOS and Linux.

LiveKit Agents plugin for tts-cache-router, a caching TTS router for Sarvam, Raya, ElevenLabs, OpenAI, Deepgram, xAI and Soniox

Audit Claude Code's prompt-cache usage from local transcripts: subagents, cache rebuilds by cause, idle compactions.

Temporary package placeholder for staged publishing

Temporary package placeholder for staged publishing