Python agents

MCP local de découverte et sélection déterministe de skills validés.

A tamper-evident, hash-chained audit ledger for AI agent actions.

Standard ACP agent over AgentHarness (no x.ai extensions)

Connect a voice agent to the TeleCMI telephony platform

Drag files found on the command line into any app that accepts drops

Fake AG-UI agent backend for testing. Deterministic scripted event streams, session record & replay, failure injection. No LLM needed.

Lightweight, hardware-independent local GGUF parser and Ollama substitute server with Hugging Face Hub integration.

Automated AI Pull Request Reviewer & Security Code Auditor.

Static dead-code analyzer for Python projects

Autonomous Security & Zero-Trust Governance Guardian for Enterprise AI Agent Fleets by Prema AI Labs

A deterministic evidence gate for coding-agent completion claims.

Permissioned coding-agent CLI with durable planning and isolated execution.

LiteLLM provider backend for purrcept_core.

Open-source policy engine for agentic payments

Packaged Carta service agent binary and utilities for pycarta[service].

MemorySync memory for Pydantic AI: a drop-in agent capability that injects recalled context every run and persists completed turns, plus typed memory tools and async helpers.

Speak pydantic-ai streamed output aloud, as it streams

Run Pydantic AI graphs with durable Waymark model and tool actions

Harness, loop and graph engineering for production AI agents.

Regression tests for AI agents. Record once, replay offline, catch behavioural regressions.

A lightweight, hackable mini-OpenCode written in Python.

Runtime-scoped LLM hooks for verified Python handlers and source patches

Gate every LangChain agent action through Qerexa runtime trust enforcement.

MCP servers extracted from the trading swarm: point-in-time data, pre-trade invariant + leakage audits, regime-fragility probes, and the Shadow Tournament bridge that scores your genome against the swarm's live champion.

Independent multi-harness, multi-model agent tools for quant_trade

Official AutoGen (v0.4+) tool wrapper for Quesen — deterministic A2A risk validation for AutoGen agents.

Official CrewAI tool wrapper for Quesen — deterministic A2A risk validation for CrewAI agents and crews.

Official LangChain / LangGraph tool wrapper for Quesen — deterministic A2A risk validation.

A Gradio RAG application combining Quivr document search with MemPalace memory

QwenPaw is a **personal assistant** that runs in your own environment. It talks to you over multiple channels (DingTalk, Feishu, QQ, Discord, iMessage, etc.) and runs scheduled tasks according to your configuration. **What it can do is driven by Skills — the possibilities are open-ended.** Built-in skills include cron, PDF/Office handling, news digest, file reading, and more; you can add custom skills. All data and tasks run on your machine; no third-party hosting.

RAG evaluation toolkit with readable LLM-judge prompts and local-first (Ollama) support.

A Professional, Enterprise-Grade Retrieval-Augmented Generation System for PDF Documents

RAG evaluation and root-cause diagnosis SDK/CLI

Health checks for RAG knowledge bases: staleness, orphans, duplicates, coverage

Evaluation harness for indirect prompt injection and retrieval poisoning resistance in RAG pipelines

A modular, fully local reliability layer for RAG pipelines.

A modular, fully local reliability layer for RAG pipelines.

Kubernetes deployment manifests for RagLeap Core, live-tested against a real cluster

Pluggable vector backends beyond ragleap-rag core's 6 (PgVector, FAISS, Pinecone, Weaviate, Qdrant, Milvus) - additional VectorBackend implementations as optional extras, no vendor lock-in, no bloated core dependency footprint.

The complete RAG evaluation library. Scores, diagnoses, and domain-aware.

Local one-shot YAML/JSON agent workflow engine + MCP toolkit (BYOK).

AI-powered reconciliation agent for finance.

High-throughput LLM Gateway and Circuit Breaker for distributed AI agents

Haystack integration for rhesis

Schema definitions and utilities for the AgentLogs dataset

LangChain integration for SafePrompt — prompt-injection detection as a callback handler. Validates every prompt flowing through a chain before it reaches the LLM.

CLI + MCP Server client for the sale-agent open platform

Official Python Telemetry & Monitoring Client for Sannex AI Operations

Academic paper knowledge-base MCP tool with ingest, retrieval and chat over your document corpus

Retrieval-augmented generation, built around your scientific domain.

OpenAI-compatible chat model for Semora: official openai SDK, LangChain chunks.

A markdown file in, a callable agent out. One Python file over the Claude Agent SDK.

Build agents with explicit steps, check them against your project, and follow a staged procedure that keeps things from drifting.

Shared SingleStore helpers for langchain-singlestore and langgraph-singlestore

Skein — team coordination platform for humans + AI agents, built on the Strands Agents SDK

Self-improving skills for Claude Code — your agent learns, recalls, reinforces, and forgets. SQLite + FTS5 + local embeddings + MCP.

Proxy inteligente entre APIs compatíveis com OpenAI e múltiplos provedores de LLM, com roteamento por complexidade e compressão de contexto.

Agent workstation capability discovery, diagnostics, and reusable document tools

Haystack Document Store for Apache Solr

Convierte montos a letras en español, inglés, portugués y francés para cheques, facturas, recibos y contratos, con soporte para las 154 divisas del estándar ISO 4217.

Stellarmesh 对象存储控制面与预签名数据面客户端

Use pipx run stem-agent to install the STEM toolkit for coding agents.

Provider-agnostic contracts for Storage Kernel.

Shared Storage Kernel implementation: paths, streams, observability and error normalization.

Dynamic provider factory for Storage Kernel.

AWS S3 provider package for Storage Kernel.

GCP Storage provider package for Storage Kernel.

Local filesystem provider for Storage Kernel.

SFTP provider package for Storage Kernel.

Run bigger LLMs on smaller GPUs through intelligent, asynchronous layer streaming.

OpenTelemetry Anthropic instrumentation

A tick-based / real-time multi-agent swarm runtime with hierarchical memory, a hybrid LLM/rule-based orchestrator, and pluggable domain physics -- for simulations and for real hardware/services alike.

Multi-methodology swarm penetration testing agent built with LangGraph

TARS — a disciplined, multi-session coding agent loop

Minimal, dependency-light Python SDK for building autonomous Ed25519 agents on the Technocore protocol

Tencent Cloud CLS observability for AgentScope agents

CLS observability SDK for OpenAI Agents SDK — per-turn trace model, direct upload to Tencent Cloud CLS

GGUF runtime, model manager & OpenAI server utilizing device resources for Android Termux & ARM64

OpenAI-compatible model, prompt, structured invocation, embedding, and memory APIs for tketool

Extensible data structures over pluggable ordered storage backends

OpenTelemetry-first LLM tracing SDK with OpenAI instrumentation

Python SDK for TollMeshCache - Distributed CRDT-based caching

LangChain integration for ToolTrace web intelligence API

LlamaIndex integration for ToolTrace web intelligence API

Pre-commit hook and CLI that fails when a branch adds no towncrier changelog fragment.

Local stdio MCP servers extracted from the trading swarm: point-in-time data assembly, pre-trade invariant + leakage auditing, and regime-fragility probing.

Claude Code hook instrumentation plugin for TruLens.

Library to systematically track and evaluate LLM based applications.

Code-first executable hooks and strict schemas for Codex and Claude Code.

A unified interface to all major LLM providers with system certificate and async support

Diversity-aware retrieval for RAG: the Vendi Retrieval Score, standalone or in the Vendi-RAG loop

Framework-neutral Python SDK for native Vintraa agents.

CLM: context-lifecycle management perception plugin for vLLM-HUST (request-lifecycle event sink)

DLA: decode-length-aware preemption & admission for vLLM-HUST

Benchmark-driven configuration optimization for vLLM inference servers

SBW watermarking logits processor for vLLM

Spark2_5 support for vLLM, maintained out of tree

LiteLLM (virtual key lifecycle + usage reporting) plugin for Waldur Site Agent

A terminal coding agent

Composable RAG primitives on Qdrant — a library, a CLI, and an MCP server