AgentScout / By role / AI engineers

AI agent tools for AI engineers (2026)

Frameworks, memory, MCP, evals and observability for engineers who build and run their own agents. We picked the top tools for 7 jobs AI engineers do every day: 37 tools in all, 34 of them free or with a free tier.

Jump to: Build agents · Multi-agent systems · Memory · Build MCP servers · Evaluate agents · Monitor in production · Deploy and host

Free tools from AgentScout for AI engineers: AI API cost calculator · Agent setup generator

How we picked: each section takes the tools matched to that task in our directory, putting tools whose category and stated use cases fit the task first, then ordering by tracked GitHub stars. No paid placements.

Build agents

Frameworks and SDKs for developers building their own agents, with tools, state and orchestration built in. All 247 tools →

Hermes Agent (Nous Research) Open source
Open-source agent harness; v0.21.7 stable shipped Oct 12, v0.22.0 RCs Oct 13-14. · ★ 253,590
DeepSeek Harness Open source
DeepSeek's open-source agent harness with a local web UI, built so that every tool, model and scheduled task is a swappable plugin; still a developer preview. · ★ 251,853
AutoGPT Source-available + cloud
One of the first autonomous agent projects, now a platform for building, deploying and running continuous AI agents that automate workflows. · ★ 187,612
Ruflo Open source
Open-source multi-agent meta-harness for Claude Code, Codex, and MCP clients; Oct 13 release adds Mods (mission workflows) and a single control plane with human approval. · ★ 74,609
MetaGPT Open source
Multi-agent framework that assigns roles like product manager and engineer to LLM agents to build software. · ★ 70,885
AutoGen Open source
Microsoft's framework for multi-agent conversations, now archived and no longer maintained; Microsoft points new users to Microsoft Agent Framework. · ★ 61,395

Multi-agent systems

Frameworks for teams of agents that hand work to each other, split tasks and check each other's output. All 78 tools →

Ruflo Open source
Open-source multi-agent meta-harness for Claude Code, Codex, and MCP clients; Oct 13 release adds Mods (mission workflows) and a single control plane with human approval. · ★ 74,609
AutoGen Open source
Microsoft's framework for multi-agent conversations, now archived and no longer maintained; Microsoft points new users to Microsoft Agent Framework. · ★ 61,395
CrewAI Open source + cloud
Framework for role-based multi-agent crews. · ★ 59,675
Agno Open source + cloud
Python framework and runtime for building, running and managing multi-agent systems with memory, knowledge and tools. · ★ 42,769
ChatDev Open source
Zero-code multi-agent platform from OpenBMB, grown from a software-development agent company into a general multi-agent orchestration system. · ★ 34,545
AgentScope Open source
Framework for building and running transparent multi-agent applications with visual debugging. · ★ 33,376

Memory

Memory layers that let agents remember users, past conversations and facts across sessions. All 97 tools →

MarkItDown Open source
Microsoft's Python tool and MCP server that converts PDFs, Office documents, images, audio and web pages to Markdown for LLM and agent pipelines. · ★ 190,824
Claude-Mem Open source
Gives coding agents memory across sessions by capturing what they do, compressing it and feeding the relevant context back in later. · ★ 101,230
RAGFlow Open source + cloud
Open-source RAG engine with deep document parsing and agent features for building grounded question-answering over your files. · ★ 92,335
Headroom Open source
Library, proxy and MCP server that compresses tool outputs, logs, files and RAG chunks locally before they reach the LLM, cutting tokens for coding agents. · ★ 75,629
Mem0 Open source + cloud
Mem0 is an open-source memory layer for AI agents that extracts facts from conversations in a single add-only pass, links entities across memories and retrieves them in later sessions, with an Apache-2.0 SDK and a managed platform. · ★ 67,147
MemPalace Open source
Local-first AI memory system that stores conversations verbatim with a pluggable backend and serves recall to agents without API calls. · ★ 59,622

Build MCP servers

Tools and SDKs for building, hosting and publishing your own MCP servers. All 34 tools →

Pi Open source
Minimal, extensible open-source terminal coding agent and agent toolkit from Earendil, with a multi-provider LLM API, TypeScript SDK, RPC mode and installable extensions, skills and themes. · ★ 116,073
LiteLLM Open source + cloud
Open-source AI gateway and Python SDK that calls 100+ LLM providers in the OpenAI format, with cost tracking, guardrails, load balancing and logging. · ★ 61,465
Kong Gateway Open source + cloud
Cloud-native API, LLM and MCP gateway that adds routing, auth, multi-LLM support, semantic security and MCP traffic analytics through plugins. · ★ 44,278
OpenAI Agents SDK (Python) Open source
OpenAI's Python SDK for building multi-agent workflows; v0.23.0/0.23.1 released Oct 1-2. · ★ 30,023
FastMCP Open source
Pythonic framework for building MCP servers and clients quickly. · ★ 28,106
MCP Python SDK Open source
Official Python SDK for building Model Context Protocol servers and clients. · ★ 24,595

Evaluate agents

Evaluation tools for scoring agent outputs, catching regressions and comparing prompts and models. All 115 tools →

MLflow Open source + cloud
Open-source platform for tracing, evaluating and managing agents, LLM apps and ML models. · ★ 28,418
Opik Open source + cloud
Comet's open-source platform for tracing, evaluating and monitoring LLM apps, RAG systems and agent workflows, self-hosted or in the cloud. · ★ 22,700
Arize Phoenix Source-available + cloud
Open-source tracing and evaluation for LLM apps. · ★ 11,828
SWE-bench Open source
Benchmark that tests whether language models and coding agents can resolve real GitHub issues, with multimodal and verified variants. · ★ 6,067
Kiln Source-available
Free desktop app and open-source library for building AI systems with evals, RAG, agents, fine-tuning and synthetic data in one workbench. · ★ 5,195
Agenta Open source + cloud
Open-source LLMOps platform for prompt management, evaluation and observability. · ★ 4,840

Monitor in production

Observability for agents in production: traces, costs, latency and errors in one place. All 34 tools →

MLflow Open source + cloud
Open-source platform for tracing, evaluating and managing agents, LLM apps and ML models. · ★ 28,418
Opik Open source + cloud
Comet's open-source platform for tracing, evaluating and monitoring LLM apps, RAG systems and agent workflows, self-hosted or in the cloud. · ★ 22,700
Arize Phoenix Source-available + cloud
Open-source tracing and evaluation for LLM apps. · ★ 11,828
OneUptime MCP Open source + cloud
OneUptime's official MCP server, giving agents access to monitors, incidents, on-call and status pages in this open-source observability platform. Free plan, paid plans from $22 per user, or self-host. · ★ 7,751
Failproof AI Open source
Observability and policy enforcement for AI agent harnesses. · ★ 5,560
agentgateway Open source
Proxy for AI agents and MCP servers with security and observability. · ★ 5,373

Deploy and host

Infrastructure for running agents in production: hosting, scaling and serverless runtimes. All 24 tools →

Sim Studio Open source + cloud
Workspace for visually building, deploying and monitoring AI agents and workflows. · ★ 29,835
Google ADK Open source
Google's Agent Development Kit for building and deploying agents. · ★ 21,826
RocketRide Open source
AI pipeline engine with a C++ core and 50+ Python nodes for building, debugging and deploying LLM workflows and agents from VS Code or the CLI. · ★ 17,753
Trigger.dev Open source + cloud
Platform for building and deploying durable background AI agents and workflows in TypeScript. · ★ 16,558
VoltAgent Open source + cloud
TypeScript agent framework plus an engineering platform for building, observing and deploying AI agents. · ★ 10,819
Nasiko Open source
Open runtime for deploying and operating AI agents. · ★ 9,479