{ "category": "AI agent harnesses", "slug": "ai-agent-harnesses", "publisher": "BuildStack.ai (AI FUTURE ASIA PTE LTD, Singapore)", "verified_on": "2026-10-20", "timezone": "Asia/Singapore", "count": 15, "notes": "Facts verified from vendor documentation, READMEs and the GitHub API on the verified_on date; each listing carries its source URLs. 'unknown' means not verified, not absent. No prices are listed. featured=true marks a paid Sponsored listing; editorial facts stay independent of payment.", "listings": [ { "id": "langgraph", "name": "LangGraph", "featured": false, "url": "https://www.langchain.com/langgraph", "type": "Open source + managed", "description": "Low-level orchestration framework from LangChain for building, managing and deploying long-running, stateful agents as graphs.", "github_url": "https://github.com/langchain-ai/langgraph", "github_stars": 43523, "github_note": "Python repo; LangGraph.js is a separate repo.", "license": "MIT", "languages": "Python, JavaScript/TypeScript (LangGraph.js)", "hosting": "Self-hosted library; LangSmith Deployment (Cloud, BYOC or self-hosted) for managed runtime", "caps": { "multi_agent": "yes", "durable": "yes", "hitl": "yes", "tracing": "yes", "mcp": "yes" }, "capability_notes": "Durable execution that resumes from where a run left off; human-in-the-loop via interrupts; short- and long-term memory; tracing and evals through LangSmith (separate product); MCP via LangChain's MCP adapter (langchain[mcp]); documented multi-agent patterns (subagents, handoffs, router).", "best_fit": "Teams that want explicit, graph-level control over long-running stateful agents and may later want a managed deployment runtime.", "sources": [ "https://github.com/langchain-ai/langgraph", "https://docs.langchain.com/oss/python/langchain/mcp", "https://docs.langchain.com/oss/python/langchain/multi-agent", "https://docs.langchain.com/langsmith/deployment", "https://api.github.com/repos/langchain-ai/langgraph" ], "github_stars_as_of": "2026-10-20" }, { "id": "openai-agents-sdk", "name": "OpenAI Agents SDK", "featured": false, "url": "https://openai.github.io/openai-agents-python/", "type": "Open source", "description": "Lightweight, provider-agnostic framework from OpenAI for multi-agent workflows built from agents, tools, handoffs, guardrails, sessions and tracing.", "github_url": "https://github.com/openai/openai-agents-python", "github_stars": 30144, "github_note": "Python repo; the JS/TS SDK (openai-agents-js) is separate.", "license": "MIT", "languages": "Python, JavaScript/TypeScript", "hosting": "Self-hosted library (runs in your process)", "caps": { "multi_agent": "yes", "durable": "yes", "hitl": "yes", "tracing": "yes", "mcp": "yes" }, "capability_notes": "Handoffs let an agent delegate to another agent; tool-approval pause/resume for human-in-the-loop; sessions (SQLAlchemy, SQLite, Redis, MongoDB, Dapr, encrypted); durable long-running runs via documented Dapr, Temporal, Restate and DBOS integrations; built-in tracing; MCP and hosted tools; sandbox, realtime and voice agents.", "best_fit": "Teams that want a small, opinionated agent loop with built-in tracing, and can add a durable-execution engine when runs get long.", "sources": [ "https://github.com/openai/openai-agents-python", "https://openai.github.io/openai-agents-python/running_agents/", "https://openai.github.io/openai-agents-python/human_in_the_loop/", "https://openai.github.io/openai-agents-python/handoffs/", "https://openai.github.io/openai-agents-python/mcp/", "https://api.github.com/repos/openai/openai-agents-python" ], "github_stars_as_of": "2026-10-20" }, { "id": "claude-agent-sdk", "name": "Claude Agent SDK", "featured": false, "url": "https://code.claude.com/docs/en/agent-sdk/overview", "type": "SDK (Anthropic Commercial Terms)", "description": "Anthropic's SDK that exposes the Claude Code agent loop, built-in tools, permissions, sessions and hooks to Python and TypeScript applications.", "github_url": "https://github.com/anthropics/claude-agent-sdk-python", "github_stars": 8300, "github_note": "Python repo; the TypeScript SDK is a separate repo.", "license": "Use governed by Anthropic Commercial Terms of Service (GitHub reports MIT for the Python repo)", "languages": "Python, TypeScript", "hosting": "Self-hosted: the SDK supervises a Claude Code subprocess in your container/sandbox; Anthropic separately offers hosted Managed Agents", "caps": { "multi_agent": "yes", "durable": "yes", "hitl": "yes", "tracing": "yes", "mcp": "yes" }, "capability_notes": "Subagents; permission modes, rules and a canUseTool callback for approvals; hooks; sessions that can be resumed, forked and persisted to external storage (resume across hosts); file checkpointing/rewind; OpenTelemetry observability; external MCP servers and in-process SDK MCP servers for custom tools.", "best_fit": "Teams building Claude-based coding or computer-use style agents who want Claude Code's tools and permission model inside their own app.", "sources": [ "https://github.com/anthropics/claude-agent-sdk-python", "https://code.claude.com/docs/en/agent-sdk/overview", "https://code.claude.com/docs/en/agent-sdk/subagents", "https://code.claude.com/docs/en/agent-sdk/permissions", "https://code.claude.com/docs/en/agent-sdk/sessions", "https://code.claude.com/docs/en/agent-sdk/hosting", "https://api.github.com/repos/anthropics/claude-agent-sdk-python" ], "github_stars_as_of": "2026-10-20" }, { "id": "google-adk", "name": "Google Agent Development Kit (ADK)", "featured": false, "url": "https://adk.dev/", "type": "Open source + managed", "description": "Google's open-source framework for building, debugging and deploying AI agents, from single agents to graph and multi-agent workflows.", "github_url": "https://github.com/google/adk-python", "github_stars": 21913, "github_note": "Python repo; Java, Go and TypeScript ADKs are separate repos.", "license": "Apache-2.0", "languages": "Python, TypeScript, Go, Java, Kotlin", "hosting": "Self-host anywhere containers run; documented deployment to Agent Runtime (Google Cloud Agent Platform), Cloud Run and GKE", "caps": { "multi_agent": "yes", "durable": "yes", "hitl": "yes", "tracing": "yes", "mcp": "yes" }, "capability_notes": "Modular multi-agent systems and graph workflows; human input and tool action confirmations; resumable agents (resume a stopped workflow); built-in logging, metrics and traces; MCP tools via McpToolset and agents exposed as MCP servers; A2A protocol.", "best_fit": "Teams on Google Cloud or needing agents in several languages (incl. Java/Go) with a managed deployment path.", "sources": [ "https://github.com/google/adk-python", "https://adk.dev/", "https://adk.dev/runtime/resume/", "https://adk.dev/observability/", "https://adk.dev/deploy/", "https://adk.dev/tools-custom/mcp-tools/", "https://api.github.com/repos/google/adk-python" ], "github_stars_as_of": "2026-10-20" }, { "id": "microsoft-agent-framework", "name": "Microsoft Agent Framework", "featured": false, "url": "https://learn.microsoft.com/en-us/agent-framework/overview/", "type": "Open source + managed", "description": "Microsoft's multi-language framework for production agents and multi-agent workflows; the stated successor to both AutoGen and Semantic Kernel.", "github_url": "https://github.com/microsoft/agent-framework", "github_stars": 14209, "github_note": "Go SDK lives in microsoft/agent-framework-go.", "license": "MIT", "languages": "Python, C#/.NET, Go", "hosting": "Self-hosted library; Foundry Hosted Agents for managed hosting on Microsoft Foundry", "caps": { "multi_agent": "yes", "durable": "yes", "hitl": "yes", "tracing": "yes", "mcp": "yes" }, "capability_notes": "Graph-based workflows with sequential, concurrent, handoff and group-collaboration patterns; checkpointing, human-in-the-loop and time-travel; agent sessions and middleware; built-in OpenTelemetry tracing; MCP clients for tools and A2A interoperability; declarative YAML agents; DevUI.", "best_fit": ".NET and Python shops, especially on Azure/Microsoft Foundry, and teams migrating from AutoGen or Semantic Kernel.", "sources": [ "https://github.com/microsoft/agent-framework", "https://learn.microsoft.com/en-us/agent-framework/overview/", "https://github.com/microsoft/autogen", "https://github.com/microsoft/semantic-kernel", "https://api.github.com/repos/microsoft/agent-framework" ], "github_stars_as_of": "2026-10-20" }, { "id": "crewai", "name": "CrewAI", "featured": false, "url": "https://www.crewai.com", "type": "Open source + managed", "description": "Python framework for multi-agent automation: Crews of role-based agents plus event-driven Flows, with CrewAI AMP as the commercial control plane.", "github_url": "https://github.com/crewAIInc/crewAI", "github_stars": 59879, "github_note": "", "license": "MIT", "languages": "Python", "hosting": "Self-hosted open-source framework; CrewAI AMP Suite for managed deployment, observability and governance", "caps": { "multi_agent": "yes", "durable": "yes", "hitl": "yes", "tracing": "yes", "mcp": "yes" }, "capability_notes": "Crews (autonomous agent teams) and Flows (event-driven workflows with managed state); checkpointing and async execution; human input / human-in-the-loop workflows; tracing and observability; MCP servers as tools and A2A support.", "best_fit": "Teams prototyping role-based multi-agent automations in Python who may want a vendor-run control plane later.", "sources": [ "https://github.com/crewAIInc/crewAI", "https://docs.crewai.com/en/concepts/flows", "https://docs.crewai.com/en/learn/human-in-the-loop", "https://docs.crewai.com/en/mcp/overview", "https://api.github.com/repos/crewAIInc/crewAI" ], "github_stars_as_of": "2026-10-20" }, { "id": "pydantic-ai", "name": "Pydantic AI", "featured": false, "url": "https://pydantic.dev/docs/ai/overview/", "type": "Open source", "description": "Python agent framework from the Pydantic team focused on type-safe agents, with a capabilities system, durable execution integrations and OpenTelemetry instrumentation.", "github_url": "https://github.com/pydantic/pydantic-ai", "github_stars": 20733, "github_note": "", "license": "MIT", "languages": "Python", "hosting": "Self-hosted library; runs inside Temporal, DBOS, Prefect and other engines for durability", "caps": { "multi_agent": "yes", "durable": "yes", "hitl": "yes", "tracing": "yes", "mcp": "yes" }, "capability_notes": "Multi-agent patterns (agent delegation, programmatic hand-off, graphs); durable execution via TemporalDurability, DBOS and Prefect integrations, plus step persistence to save/resume/fork runs; deferred tools for human-in-the-loop tool approval; OpenTelemetry-native instrumentation (Pydantic Logfire or any OTel backend); MCP as a core capability.", "best_fit": "Python teams that want typed, testable agents and to pick their own durable-execution engine.", "sources": [ "https://github.com/pydantic/pydantic-ai", "https://pydantic.dev/docs/ai/tools-toolsets/deferred-tools/", "https://pydantic.dev/docs/ai/guides/multi-agent-applications/", "https://api.github.com/repos/pydantic/pydantic-ai" ], "github_stars_as_of": "2026-10-20" }, { "id": "mastra", "name": "Mastra", "featured": false, "url": "https://mastra.ai", "type": "Open source + managed", "description": "TypeScript framework for AI applications and agents, with workflows, memory, evals, observability and a local Studio.", "github_url": "https://github.com/mastra-ai/mastra", "github_stars": 28917, "github_note": "", "license": "Apache-2.0 core; code in ee/ directories under the Mastra Enterprise License (GitHub reports NOASSERTION)", "languages": "TypeScript (Node.js, Bun, Deno, Cloudflare)", "hosting": "Self-host on any Node.js-compatible environment (Mastra server, web framework or cloud provider); Mastra platform for observability, Studio and serving", "caps": { "multi_agent": "yes", "durable": "yes", "hitl": "yes", "tracing": "yes", "mcp": "yes" }, "capability_notes": "Supervisor agents that coordinate subagents; workflows with snapshots, suspend/resume, time travel and human-in-the-loop; traces, metrics and evals; MCPClient to consume MCP servers and MCPServer to expose agents, tools and workflows.", "best_fit": "TypeScript/JavaScript teams who want agents and workflows in the same stack as their web app.", "sources": [ "https://github.com/mastra-ai/mastra", "https://mastra.ai/docs", "https://mastra.ai/docs/subagents", "https://mastra.ai/docs/workflows/suspend-and-resume", "https://mastra.ai/docs/observability/overview", "https://mastra.ai/docs/connections/mcp", "https://mastra.ai/docs/deployment/overview", "https://raw.githubusercontent.com/mastra-ai/mastra/main/LICENSE.md", "https://api.github.com/repos/mastra-ai/mastra" ], "github_stars_as_of": "2026-10-20" }, { "id": "vercel-ai-sdk", "name": "Vercel AI SDK", "featured": false, "url": "https://ai-sdk.dev", "type": "Open source", "description": "Provider-agnostic TypeScript toolkit from Vercel for AI applications and agents, with tool-loop agents, subagents and durable workflow agents.", "github_url": "https://github.com/vercel/ai", "github_stars": 27400, "github_note": "", "license": "Apache-2.0 (per LICENSE file; GitHub reports NOASSERTION)", "languages": "TypeScript (Node.js; React, Next.js, Svelte, Vue, Angular UI integrations)", "hosting": "Self-hosted library", "caps": { "multi_agent": "yes", "durable": "yes", "hitl": "yes", "tracing": "yes", "mcp": "yes" }, "capability_notes": "ToolLoopAgent and subagents a parent can delegate to; WorkflowAgent (@ai-sdk/workflow) for durable, resumable agents with automatic state persistence; tool execution approval (toolApproval); OpenTelemetry telemetry via @ai-sdk/otel; MCP tools.", "best_fit": "TypeScript/Next.js teams adding agents to web products.", "sources": [ "https://github.com/vercel/ai", "https://ai-sdk.dev/docs/agents/overview", "https://ai-sdk.dev/docs/agents/subagents", "https://ai-sdk.dev/docs/agents/workflow-agent", "https://ai-sdk.dev/docs/ai-sdk-core/tools-and-tool-calling", "https://ai-sdk.dev/docs/ai-sdk-core/telemetry", "https://ai-sdk.dev/docs/ai-sdk-core/mcp-tools", "https://raw.githubusercontent.com/vercel/ai/main/LICENSE", "https://api.github.com/repos/vercel/ai" ], "github_stars_as_of": "2026-10-20" }, { "id": "strands-agents", "name": "Strands Agents", "featured": false, "url": "https://strandsagents.com", "type": "Open source", "description": "Open-source SDK and ready-made agent harness for Python and TypeScript that runs in-process with no hosted control plane; Amazon Bedrock is the default model provider.", "github_url": "https://github.com/strands-agents/harness-sdk", "github_stars": 8866, "github_note": "Monorepo (formerly strands-agents/sdk-python).", "license": "Apache-2.0", "languages": "Python, TypeScript", "hosting": "Self-hosted (in-process); documented deployment to AWS Lambda, EKS, Fargate, EC2 or Amazon Bedrock AgentCore Runtime", "caps": { "multi_agent": "yes", "durable": "yes", "hitl": "yes", "tracing": "yes", "mcp": "yes" }, "capability_notes": "Multi-agent handoffs, swarms, graphs and workflows; sessions that persist and resume across restarts plus long-term memory; human approval gates for tool calls (interventions); tracing and evals; MCP tools; model-agnostic (Bedrock, Anthropic, OpenAI, Gemini, Ollama and more).", "best_fit": "AWS-centric teams that want an in-process agent harness with a serverless deployment path.", "sources": [ "https://github.com/strands-agents/harness-sdk", "https://strandsagents.com/llms.txt", "https://api.github.com/repos/strands-agents/harness-sdk" ], "github_stars_as_of": "2026-10-20" }, { "id": "agno", "name": "Agno", "featured": false, "url": "https://www.agno.com", "type": "Open source", "description": "Python framework and runtime (AgentOS) for running agents, teams and workflows as a service you host, with a web control plane.", "github_url": "https://github.com/agno-agi/agno", "github_stars": 42814, "github_note": "", "license": "Apache-2.0", "languages": "Python", "hosting": "Self-hosted AgentOS runtime on any container platform (Docker, Railway, AWS, GCP); data stored in your own database", "caps": { "multi_agent": "yes", "durable": "yes", "hitl": "yes", "tracing": "yes", "mcp": "yes" }, "capability_notes": "Teams of agents (incl. nested teams) and workflows; sessions, memory and traces stored in your database; human-in-the-loop (user confirmation, user input, external execution, admin approval); OpenTelemetry tracing and audit logs; exposes agents via a REST API and MCP server; MCP context providers. By default sends a telemetry event per agent run (vendor states prompts, messages and outputs are never sent); AGNO_TELEMETRY=false disables it.", "best_fit": "Teams that want to run a self-hosted multi-tenant agent platform with an API and control plane.", "sources": [ "https://github.com/agno-agi/agno", "https://docs.agno.com/", "https://docs.agno.com/teams/overview", "https://docs.agno.com/hitl/overview", "https://docs.agno.com/features/observability", "https://api.github.com/repos/agno-agi/agno" ], "github_stars_as_of": "2026-10-20" }, { "id": "smolagents", "name": "smolagents", "featured": false, "url": "https://huggingface.co/docs/smolagents/index", "type": "Open source", "description": "Hugging Face's minimal Python library for agents that write and execute their actions as code, with tool-calling agents also supported.", "github_url": "https://github.com/huggingface/smolagents", "github_stars": 29929, "github_note": "", "license": "Apache-2.0", "languages": "Python", "hosting": "Self-hosted library; optional remote/sandboxed code execution", "caps": { "multi_agent": "yes", "durable": "unknown", "hitl": "unknown", "tracing": "yes", "mcp": "yes" }, "capability_notes": "CodeAgent and ToolCallingAgent; multi-agent hierarchies (managed agents); tools from any MCP server, LangChain or the Hugging Face Hub; share agents to the Hub; run inspection with OpenTelemetry; model-agnostic. Durable state and human-in-the-loop not verified.", "best_fit": "Developers and researchers who want a small, readable code-acting agent loop.", "sources": [ "https://github.com/huggingface/smolagents", "https://huggingface.co/docs/smolagents/tutorials/inspect_runs", "https://huggingface.co/docs/smolagents/examples/multiagents", "https://api.github.com/repos/huggingface/smolagents" ], "github_stars_as_of": "2026-10-20" }, { "id": "letta", "name": "Letta (Letta Code harness)", "featured": false, "url": "https://www.letta.com", "type": "Open source + managed", "description": "Stateful agent harness (formerly MemGPT) for agents with persistent memory, identity and skills that learn over time; usable from a CLI, the web, messaging channels or the Letta Agent SDK.", "github_url": "https://github.com/letta-ai/letta-code", "github_stars": 3657, "github_note": "Current source repo. The original letta-ai/letta repo (25,193 stars) now points here and keeps the retired V1 server on an archive branch.", "license": "Apache-2.0", "languages": "CLI via npm/uv/pipx; Letta Agent SDK for TypeScript", "hosting": "Letta Cloud (default) or local / self-hosted App Server", "caps": { "multi_agent": "yes", "durable": "yes", "hitl": "yes", "tracing": "unknown", "mcp": "yes" }, "capability_notes": "Persistent stateful agents with git-versioned memory (MemFS), memory blocks and skill learning; subagents and agents calling other agents; permission modes with interactive tool approvals; schedules/crons; hosted Letta MCP server (Letta Cloud) lets MCP clients create and message Letta agents. Tracing/observability not verified.", "best_fit": "Teams building long-lived personal or shared agents where persistent memory is the core requirement.", "sources": [ "https://github.com/letta-ai/letta", "https://github.com/letta-ai/letta-code", "https://docs.letta.com/llms.txt", "https://api.github.com/repos/letta-ai/letta-code" ], "github_stars_as_of": "2026-10-20" }, { "id": "temporal", "name": "Temporal", "featured": false, "url": "https://temporal.io/solutions/ai", "type": "Open source + managed", "description": "Durable execution platform used as the orchestration layer under AI applications and agents; not an agent framework itself, but integrates with several.", "github_url": "https://github.com/temporalio/temporal", "github_stars": 23833, "github_note": "Stars are for the Temporal server.", "license": "MIT", "languages": "Go, Java, Python, TypeScript, .NET, Ruby, PHP, Rust (official SDKs)", "hosting": "Self-hosted open-source server or Temporal Cloud (managed)", "caps": { "multi_agent": "unknown", "durable": "yes", "hitl": "yes", "tracing": "yes", "mcp": "unknown" }, "capability_notes": "Durable Execution: workflows hold state over long periods and recover from failures; human-in-the-loop (e.g. approving agent decisions); OpenTelemetry tracing of workflows and activities (e.g. Python SDK OpenTelemetryPlugin); listed integrations with OpenAI Agents SDK, Google ADK, Mastra and others; AI Cookbook. Multi-agent primitives and MCP support not verified.", "best_fit": "Platform teams that need agent runs to survive crashes, deploys and multi-day waits, alongside an agent framework.", "sources": [ "https://github.com/temporalio/temporal", "https://temporal.io/solutions/ai", "https://docs.temporal.io/encyclopedia/architecture/temporal-sdks", "https://docs.temporal.io/develop/python/platform/observability", "https://docs.temporal.io/cloud", "https://api.github.com/repos/temporalio/temporal" ], "github_stars_as_of": "2026-10-20" }, { "id": "inngest-agentkit", "name": "Inngest AgentKit", "featured": false, "url": "https://agentkit.inngest.com", "type": "Open source + managed", "description": "TypeScript library for multi-agent networks with deterministic routing and MCP tooling, which gains retries, concurrency control and traces when run on Inngest.", "github_url": "https://github.com/inngest/agent-kit", "github_stars": 949, "github_note": "Last push to the repo per GitHub API: 2026-04-29.", "license": "Apache-2.0", "languages": "TypeScript", "hosting": "Self-hosted library served over HTTP; Inngest Platform for orchestration (Inngest Dev Server locally)", "caps": { "multi_agent": "yes", "durable": "yes", "hitl": "yes", "tracing": "yes", "mcp": "yes" }, "capability_notes": "Agents combined into Networks with a Router and shared Network State; MCP servers as tools; human-in-the-loop via Inngest waitForEvent(); with Inngest: automatic retries of agent and tool calls (default 4 with exponential backoff), multitenancy/concurrency limits and step-by-step traces. Durability depends on deploying with Inngest.", "best_fit": "TypeScript teams already on Inngest, or wanting event-driven, retryable agent networks.", "sources": [ "https://github.com/inngest/agent-kit", "https://agentkit.inngest.com/", "https://agentkit.inngest.com/concepts/deployment", "https://agentkit.inngest.com/advanced-patterns/retries", "https://agentkit.inngest.com/advanced-patterns/human-in-the-loop", "https://api.github.com/repos/inngest/agent-kit" ], "github_stars_as_of": "2026-10-20" } ] }