PyPI agents

Add crawling capability to pydantic ai agent

A standalone RAG package for Pydantic AI with support for multiple vector stores.

Add tts and stt capability to pydantic ai agent

Agentic AI audits for tool-use, memory, permissions, and prompt injection

LLM and RAG audits including safety, security, hallucination, and citation checks

Drop-in observability and trust layer for AI pipelines and agents.

SIMTOOREAL — wrap any robot training command with full end-to-end analytics.

Runtime token proxy + optimization toolkit for LLM developers and enterprises. Intercepts API calls, strips waste in real-time, tracks costs, and serves a web dashboard.

AWS CDK construct for deploying production-ready Bedrock AI agents with Lambda, DynamoDB, and OpenSearch

Placeholder for the SwarmPilot distributed task scheduling system.

Drop-in, compliance-instrumented wrapper around the official Anthropic Python SDK.

Drop-in, compliance-instrumented wrapper around the official OpenAI Python SDK.

A btop-style terminal UI for monitoring a vLLM instance and its GPU in real time.

Name reservation placeholder.

Guard LLM tool calls with rules, scoring, and audit trails.

Agent MCP Router MCP — one router for the whole MEOK fleet. Holds 62 MEOK MCPs behind one namespaced endpoint. Install ONE thing, get all of them. By MEOK AI Labs.

Agent Replay Debugger MCP — record every agent step + replay deterministically. Step-debugger for agentic loops. Audit-evidence signed exports. By MEOK AI Labs.

Agent Token Budget MCP — per-session spend cap with signed budget-exhausted attestations. Twin of bft-progress-council-mcp (spend-axis vs stall-axis guardrail). By MEOK AI Labs.

Generador de proyectos Scratch 3.0 con IA — CLI y servidor MCP

Unit tests for agents. Define an agent with capa's capabilities.yaml, write scenarios and assertions, mock MCP servers, run in CI.

Approve or deny your AI coding agent's risky commands from your phone, with an audit trail. Works with Claude Code and OpenAI Codex.

Low-code framework for generating agentic AI systems from a prompt or spec

An AI-powered Python library for context-aware data cleaning using local LLMs

AutoGen integration for the Scavio API -- 188 real-time data tools across 31 platforms (Google, Amazon, Walmart, eBay, Target, Home Depot, Zillow, Redfin, Booking, Airbnb, Tripadvisor, Yelp, Indeed, Glassdoor, G2, Capterra, App Store, Google Play, SEC EDGAR, Companies House, Google Ads, Meta Ad Library, YouTube, TikTok, TikTok Shop, Instagram, Threads, Kuaishou, X, LinkedIn, Reddit) plus extract, which reads any URL as clean Markdown

Drop-in Anthropic & OpenAI Python SDK clients for BlockRun — genuine native passthrough (real signatures, system_fingerprint, native structure) with x402 USDC micropayments on Base or Solana. No model substitution, no response reshaping. Plus AI generation clients: image, search (Grok/Exa), audio (TTS/music), and AI voice calls.

PreCompact ledger and per-component context profiler for Claude Code

Detect and revive Claude Code remote-control connections

Session checkpoint and recovery layer for Claude Code — resume interrupted sessions, export shareable HTML traces.

Arc Gate runtime governance for CrewAI agents

CrewAI integration for Scavio Search API -- 188 real-time search tools for AI agents across Google, Amazon, Walmart, eBay, Target, Home Depot, YouTube, Reddit, TikTok, Instagram, X, LinkedIn, Threads, Kuaishou, Zillow, Redfin, Booking.com, Airbnb, Tripadvisor, Yelp, Indeed, Glassdoor, the App Store, Google Play, SEC EDGAR, Companies House, G2, Capterra, Google Ads and the Meta Ad Library

An Entari Plugin for LLM Chat with Function Call

ERPNext and Frappe REST/RPC MCP Server for Agentic AI!

Um cliente, todos os LLMs. Roteamento por intencao, fallback automatico e tracking de custo.

Async No-KYC Fragment.com client. Buys Stars / Premium / Ads top-up via a remote backend without KYC.

Unified CLI for the six-tool agent stack (Node.js tool, installable via pip)

Dead-simple Groq LLM chaining in Python. Chain prompts with .step() — no LangChain needed.

Native LangChain chat models for OpenAI Codex subscription OAuth and OpenCode Zen/Go — no Node sidecar.

LLM client layer for the Lexigram Framework — OpenAI, Anthropic, Ollama, Cohere, Groq, Mistral

LlamaIndex tool spec for the Ejentum Reasoning Harness. Wraps the hosted Ejentum MCP server and exposes eight cognitive harness tools as LlamaIndex FunctionTool objects: four dynamic (reasoning, code, anti-deception, memory) plus four adaptive variants (adaptive-reasoning, adaptive-code, adaptive-anti-deception, adaptive-memory) that pre-fit the cognitive operation to the caller's task via an adapter LLM.

Route prompts to the right LLM by complexity. Supports Groq, OpenAI, Anthropic. Save cost, keep quality.

Lightweight evaluation framework that unifies inference through a single VLLM sampler and runs IF-EVAL, IFBench, WritingBench, HealthBench, Arena-Hard, and AlpacaEval end-to-end.

Local-first, CI-friendly regression testing for LLM outputs.

Decorator-based LLM-native typed Python functions — declare the signature and docstring; the LLM runs the body.

Static preflight scanner for MCP servers and agent skills

Public-safe MVR API preflight wrapper for AI agents.

loading...

Standalone observed-fact graph substrate for code and document structure

Point your agent at your docs and your RAG app; get a golden test set + an LLM-as-judge & retrieval scorecard, in one command.

Accountability infrastructure for AI agents

Enterprise-grade Python RAG stack — loaders, cleaning, chunking, embedding, and vector store in one composable library

Production-ready RAG with near-zero hallucination. Pluggable providers (OpenAI, Anthropic, Gemini, OpenRouter, Azure, Voyage, ...), graph + agentic + multilingual strategies, a FastAPI server with a bundled React UI, and one-click publishing of a configured setup as a public ask-only assistant.

TraceLLM — Open-source LLM observability, tracing, and replay infrastructure.

Document-grounded Q&A with sentence-level citations and faithfulness verification

A minimal terminal screensaver for when AI agents are working

Circuit breaker SDK for LLM agents.

AgentCamp SDK — where agents train their own experts. Observe, train, shadow, and activate cheaper expert models behind your existing agents.

An agentic middleware for automated JSON validation and self-correction from LLM outputs.

An MCP which adds terminal capabilities to an agent.

Official Python SDK for AgenticDome security guardrails, DLP, tool authorization, and multi-agent delegation enforcement.

Universal FastAPI and FastMCP gateway backed by a local Pi AgentSession

A minimal, reusable Python framework for building LLM agents with tools and an observe-act loop.

Agent orchestrator for terminal-based coding agents — discovers installed CLIs, runs them in parallel, ships giant projects in hours

Graph control for AI-agent workflows.

AI agent and MCP server security scanner — discovery, static analysis, supply chain audit, and multi-agent trust analysis

Stateful coordination layer for multi-agent AI systems

Autonomous coding copilot

Console logging utilities for Autourgos agents and orchestrators.

Optimization config loader for Azure AI Hosted Agents

Store, retrieve, and transform AI agent memories backed by Azure Cosmos DB

Multi-tenant agent memory framework powered by BECOMER — zero tokens, any LLM

An always-present chat-agent sidebar in the Django admin, powered by django-ag-ui.

Production-grade read/write split RAG retrieval system

Eight-layer middleware guardrail pipeline for LLM-powered personal finance applications

Wrap any AI agent in a rootless Podman, Docker, or Kubernetes container for automatic isolation.

Normalize LangChain, MCP, and multimodal content blocks into provider-ready text and image payloads.

Provider-agnostic persuasion guard built with LangChain and LangGraph.

LangChain-based read-only query layer for LightRAG knowledge graph

Local auth surface manager for LLM API keys and OAuth tokens.

Add your description here

A lightweight LLM inference engine supporting OpenAI-compatible APIs.

Eval gates with confidence intervals: deterministic gates, calibrated LLM-as-judge, agentic-trace evals, and a statistically honest regression gate for LLM pipelines

A control layer for LLM-in-the-loop reliability: wrap an unreliable generator in a verify-and-refuse feedback loop with a pluggable controller (deterministic checks, a model critic, or compositions) and composable 'circuits'. Keep one exact element in the loop and outputs are checked, corrected, or honestly refused. Zero runtime dependencies; runs with no model at all.

Declarative slot-based onboarding schemas for LLM-driven conversations.

Universal Python LLM Provider Abstraction Library

Real-time music generation models.

Python SDK for mink — sandboxed AI coding agent

Redis Agent Memory integration for NVIDIA NeMo Agent Toolkit

Production helpers for running Ollama under concurrent load across local and remote endpoints.

OpenAI-compatible client + worker that bridge inference requests over a Redis queue (Redis Streams), for running LLMs across heavily restricted networks.

AgenticDome security middleware and firewall plugin for OpenClaw.

Connect your OpenAI agents to Oversee — two lines of code for full observability

Python bindings for ovstorage

Image tiles to vectors to FAISS index. Three independent CLI tools.

Build searchable FAISS indexes from any document source