PyPI agents

Inter-agent message compression and token budget management for multi-agent LLM systems

18 reusable multi-agent orchestration patterns for LLMs

Difficulty-aware routing and model selection for multi-agent LLM workflows

Pattern-aware OpenTelemetry tracing for multi-agent LLM systems

Vendor-neutral RAG + LLM serving infrastructure: swappable LLM protocol and vector store (FAISS/NumPy/Qdrant), cached embedding index, and observability.

Policy-as-code guardrail enforcement for enterprise LLM applications

Python SDK and drop-in LLM guardrail for the SentraGuard backend (OpenAI / Anthropic / Gemini compatible).

Seren: A Self-Refining Decoder-Only LLM with GQA, RoPE, Flash Attention, and a novel two-pass self-refinement mechanism.

Free AI LLM client with multi-provider fallback and rate limiting

539 lazy-loading AI skills for Claude, Gemini, and Scientific agents

LangChain and LangGraph observability for Tuner SDKs

Pre-spend x402 counterparty reputation from observed behavior, with free preflight, signed V6 receipts, and a 24-tool MCP.

Handwritten ReAct Agent with TUI — Tools + MCP + Skills，agent frame

Open-source Python framework for building autonomous AI trading agents on GitHub repository momentum (paper mode).

Real-time terminal monitor for Claude Code token usage — with a clean centered dashboard, keyword analytics, and a one-line installer

Agent-first Python CLI harness for xiaokeer solution projects.

A lightweight Python framework for portable, versioned, reusable LLM extraction tasks.

AACP coordination layer for CrewAI multi-agent workflows

AACP coordination layer for LangChain multi-agent workflows

Scan a directory for AI-assistant skills, rules and subagents and copy selected ones into a chosen assistant's folder layout.

Agent Hub MCP — multi-agent task coordination with MCP transport

Self-learning memory infrastructure for AI products

Transparent feedback compression middleware for LLM coding agents.

从 Python CLI 包中安装内置 agent skill。

A pluggable protective harness for conversational AI agents — drop-in OpenAI-compatible proxy that defends against jailbreaks, prompt injection, data exfiltration, and denial-of-wallet.

MCP server for Agentberg — AI trading agents, learning from each other

Universal Memory-Import Tool — ingest user-memory from major LLM platforms into an AgentBrain workspace.

Read-only Inspector and evidence viewer package for AgentLedger

Local-first Harness Operating System for defining, evaluating, evolving, versioning, and deploying agent harnesses.

TRACE v0.2 — hardware-attested governance records for AI agents

Make your AI agents leaner, faster, and cheaper — smart context management and token compression

Query (and build) the AIND dynamic-foraging behavioral parquet database.

LangChain integration for AIPP (AI Payment Protocol)

AlgoVoi payment guardrail capability and tools for AutoGen / ag2

AlgoVoi payment and compliance plugin for Semantic Kernel

Fully automatic censorship removal for language models

Fully automatic censorship removal for language models

Standalone agent runtime for Databricks Apps — MCP, A2A, /invocations, tool routing

Framework-agnostic tool library for software architecture analysis

Spec + reference implementation for vault-native, multi-runtime AI agents

Multisite CMS with htmx, plugins, and SEO baked in.

Link's-Awakening-flavoured theme for the bragi CMS, used by zelda.eelcowesemann.nl

Lightweight self-hosted Telegram manager for Claude Code CLI

Personal claude.ai conversation archive — ingest, embed, search, and MCP server

Drop-in, MCP-native working memory for AI agents: gameplans, phases, a dependency graph, and post-hoc cascade — over plain markdown.

Context-driven scheduled tasks for LLMs with persistent memory — inspired by OpenClaw's Heartbeat.

Human-reviewed Agent DAG framework

DPX settlement and intelligence tools for LangChain agents — quote, ESG scoring, oracle status, and settlement execution.

Build LangGraph graphs from JSON configuration

A red-team benchmark for AI agents in financial workflows: measures regulatory-control bypass (sanctions evasion, threshold structuring, dual-approval defeat) and the efficacy of guardrails that try to stop it.

Proxy Gemini (Vertex AI) completions wrapped in OpenTelemetry trace spans; returns the answer plus the trace/span id.

Halo adapter for Claude on two runtimes (Python): the Agent SDK (install_halo — a PostToolUse encode hook + a single in-process halo_fetch MCP tool) and the raw Messages API (halo_format_claude.raw — create_raw_halo, no Agent SDK needed). Tool results become a shape map kept out of context.

ovos-persona (LLM/solver) agent protocol plugin for HiveMind-core

Haystack integration for Hugging Face API

Claude usage data provider for `mngr usage` (per-agent statusline shim that captures rate-limits + cost + session_id)

A JupyterLab extension with an AI-powered data analysis agent sidebar Panel.

Agent Memory via Google Keep MCP

An integration package connecting Duel Agents and LangChain

Seven LangChain tools for the FTIR.fun spectroscopy API

LangChain integration for Talor SERP API — 33 search engines in one tool

llama-index llms duel agents integration

A configurable local chatbot library with lightweight memory indexing.

Local-first AI memory cartridge: persistent, cross-session memory for Claude and any LLM via MCP

Reduce LLM API costs by up to 90% — prompt caching, smart routing, batch processing, and more

brave Search API tool for Simon Willison's LLM

Context compression sidecar for LLM applications. Reduce token costs 3-5x before calling any AI API.

Crawl any website and generate llms.txt - the AI-ready site map standard.

A fast, safe macOS CLI tool to reclaim disk space

MagAgent — CLI AI coding agent with persistent MagGraph memory. Like the Magpie: intelligent, tool-using, and never forgets.

Official Python SDK for Make Your Agent (MYA) — build AI agents with knowledge bases, tool execution, streaming chat, and per-session token usage

Managed Deep Agents — the define_deep_agent authoring interface plus the CLI that compiles and deploys a code-first Deep Agent repository to a managed LangGraph runtime.

RAG implementation pipeline with GPU support

Platform-specific compiled core extensions for myrm-agent-harness

Platform-specific compiled core extensions for myrm-agent-harness

Enterprise-ready, SaaS-native Python multi-agent framework

A thin, opinionated, local-first structured-output + logging layer over LiteLLM

Neo4j graph-store backend plugin for the Orchid AI framework

Privacy-Preserving Phragmén Voting via Secure Multi-Party Computation

CLI + library to audit and benchmark RAG pipelines

Lighthouse for RAG systems - diagnose and fix your retrieval pipeline

Generic async LLM gateway with concurrency, retries, and JSON helpers

Out-of-tree vLLM KVConnector for SemBlend semantic KV donor discovery

One unified API for OpenAI, Anthropic, Google, DeepSeek, and xAI with simple fiat billing.

Deterministic governance for Haystack pipelines and agents — policy enforcement, PII detection, cost tracking, and structured audit evidence. Supports Haystack 3.0 Agent Hooks (before_tool). No LLM in the governance path.

Keyword-driven paper search assistant that exports slides, summaries, and BibTeX.

Toroidal semantic graph retrieval for measurable multi-hop RAG trade-offs.

MiniMax plugin for vision agents

Qdrant RAG integration for Vision Agents with hybrid search

Small, reviewable, validation-gated agent skills for Codex-style project work.

Deterministic agent test recorder and replayer. Record live runs, replay as mocks. Zero dependencies.

Python SDK and CLI for the AgentGraph trust and identity platform

An MCP which adds terminal capabilities to an agent.

Generate and safely update AGENTS.md and RUNBOOK.md for coding agents

AgentTrace Python SDK - Drop-in tracing for any AI agent

FastAPI middleware for AgentVisa AI agent verification

LedgerProof adapter for the Anthropic Python SDK. Side-channel transparency receipts for EU AI Act Article 50.

Solo Founder OS agent #4 — keep EN/ZH locale JSON in sync via Claude + HITL markdown review.

Streaming JSONL cleaner for LLM fine-tuning datasets.

Python SDK for controlling visible ChatGPT Chat and Work through a local backend.

Maintains information quality for agentic workflows — assesses the freshness, provenance, fidelity, and task-specific validity of claims, then refreshes or repairs degraded ones before agents consume them.