MCP y llamadas a herramientas
atmosphere/atmosphere
Atmosphere is a Java-based real-time framework for building, streaming, and governing AI agents. It provides a unified broadcaster pipeline for WebSocket, SSE, long-polling, and gRPC, alongside built-in governance, human-in-the-loop workflows, and multi-protocol support including MCP, A2A, and AG-UI.
★ 3,8K⑂ 760Java
Apache-2.0Q98
Agentes y multiagente
langwatch/langwatch
An open-core platform for evaluating, testing, tracing, and monitoring LLM applications and AI agents before release and in production.
★ 3,5K⑂ 345TypeScript
Apache-2.0Q98
Agentes y multiagente
truera/trulens
TruLens is an open-source, OpenTelemetry-native evaluation and tracking library for LLM applications and AI agents, enabling developers to trace every step, score performance with LLM judges, and compare app versions.
★ 3,5K⑂ 319Python
MITQ98
Agentes y multiagente
agentscope-ai/reme
ReMe is a local-first memory layer for AI agents that transforms conversations and resources into readable, editable, and searchable Markdown memory files.
★ 3,3K⑂ 283Python
Apache-2.0Q98
MCP y llamadas a herramientas
neomjs/neo
A self-evolving software organism that combines a multi-threaded frontend application engine with an end-to-end AI engineering team operating via cross-model swarm coordination, persistent memory, and runtime application embodiment.
★ 3,3K⑂ 231JavaScript
MITQ98
RAG y sistemas de conocimiento
modelscope/evalscope
A streamlined and customizable framework for efficient large model (LLM, VLM, AIGC) evaluation and performance benchmarking.
★ 3,2K⑂ 440Python
Apache-2.0Q98
Inferencia, implementación y tiempo de ejecución
vllm-project/vllm-ascend
A community-maintained hardware plugin that enables vLLM to run model inference and serving workloads on supported Ascend NPUs.
★ 2,7K⑂ 2,1KC++
Apache-2.0Q98
LLM y modelos fundacionales
bionic-gpt/bionic-gpt
Bionic is a Rust-based, on-premise alternative to ChatGPT designed for private generative AI deployments. It provides a familiar chat interface, local or remote model access, team controls, retrieval-augmented assistants, data integrations, observability, and Kubernetes-oriented scaling.
★ 2,4K⑂ 239Rust
NOASSERTIONQ98
Agentes y multiagente
dstackai/dstack
A vendor-agnostic control plane for provisioning and orchestrating training, inference, development, and agentic workloads across GPU clouds, Kubernetes, and on-premises infrastructure.
★ 2,2K⑂ 248Python
MPL-2.0Q98
Inferencia, implementación y tiempo de ejecución
mozilla-ai/any-llm
any-llm is a Python SDK for communicating with multiple LLM providers through one interface. It supports switching providers and models with minimal code changes while using official provider SDKs when available.
★ 2,2K⑂ 211Python
Apache-2.0Q98
Agentes y multiagente
makecindy/cindy
Cindy is an open-source desktop and mobile AI-agent client that works with local files, signed-in applications, browsers, computers and phones. It supports Claude Code and Codex harnesses, multiple model-access options, persistent workspace context, automation and MCP integrations.
★ 2,1K⑂ 282TypeScript
Apache-2.0Q98
MCP y llamadas a herramientas
alpic-ai/skybridge
Skybridge is an open-source, full-stack TypeScript and React framework for building type-safe MCP Apps, MCP servers, and ChatGPT Apps across supported UI-enabled MCP clients.
★ 2K⑂ 133TypeScript
MITQ98
