An index and topic collection covering AI agents, agent frameworks, and agent runtimes that enable autonomous reasoning, tool use, and multi-step task execution. This profile catalogs the platforms and open-source projects that let developers build, orchestrate, and deploy LLM-powered agents, including foundational frameworks like LangChain, LangGraph, CrewAI, AutoGen, and AutoGPT, hosted agent platforms like Lindy, Relevance AI, and Composio, and provider-native agent runtimes from OpenAI, Anthropic, Google, and Microsoft. Distinct from foundation models and from Agent Skills (capability packages), the Agents topic focuses on the orchestration layer where planning, memory, tool calling, and execution come together.
Agents is profiled on the APIs.io network. Tagged areas include AI Agents, Agent Frameworks, Agent Orchestration, Agent Runtimes, and LLM Orchestration.
The Agents catalog on APIs.io includes 1 JSON-LD context.
Agents’ developer surface includes developer portal and 1 more developer resources.
Regulatory Posture applies to this provider. Its tags matched the
Horizontal (data, software, accessibility, platform) regime, so
Regulatory Posture carries 15 points of the composite.
If this regime is wrong for your business, say so on your
provider repo — the
applicability map is public and we will correct it.
Create-or-Update Ergonomics could not be measured. We hold no machine-readable contract for
this provider to read, so there is nothing to measure a write surface against. Excluded rather than scored zero:
never-measured and measured-empty are different facts. Publishing an OpenAPI is what makes this facet — and
several others — scorable at all.
The six quality facets above are damped to 85 points between them,
because the conditional facet above carries the other
15. That is why each facet's contribution is shown against a damped
maximum: raising a quality facet moves the composite by 85% of its nominal
weight, not 100%. The full arithmetic is at apis.io/rating/.
Improve this rating by publishing the missing artifacts — every area above can be raised, and the full rubric is at apis.io/rating/. Every facet and dimension name above is a link: it opens that measurement's own page — what it means, the exact checks that feed it, how the whole catalog distributes on it, and the providers at the top of it. This rating is computed from github.com/api-evangelist/agents: open an issue to ask a question, or submit a pull request to add artifacts.
Submit an artifact on GitHub — free →Manage your own listing — the Influence plan, $499/mo →
Agent frameworks like LangGraph, AutoGen, and CrewAI provide structured loops for planning, reflection, and multi-step reasoning over user goals, breaking complex tasks into ord...
Tool Calling and Function Use
Agents invoke external functions, APIs, and aggregated tool catalogs through provider-native tool calling, expanding their reach beyond text generation into real-world action.
Memory and State Management
Frameworks like Letta and LangGraph provide persistent memory, conversation state, and long-running execution context so agents can resume work, remember users, and accumulate k...
Multi-Agent Orchestration
CrewAI, AutoGen, and LangGraph enable multiple specialized agents to collaborate, delegate sub-tasks, and reach consensus, modeling teams of role-based workers around a shared g...
Provider-Native Agent Runtimes
OpenAI Assistants, Anthropic Claude tool use, Google ADK, Amazon Bedrock Agents, and Microsoft Azure AI Foundry expose first-party agent runtimes that handle planning, tool invo...
Browser and Web Agents
Agents like AgentQL and browser-driving frameworks let LLMs navigate real web pages, fill forms, and extract structured data, extending agent reach to systems that lack APIs.
Observability and Evaluation
Platforms like Portkey, TrueFoundry, and Bifrost provide tracing, evaluation, cost tracking, and gateway routing across agent runs to make agent behavior observable and debuggable.
Open Governance and Interoperability
The Linux Foundation Agentic AI Foundation and projects like kagent and AgentGateway are establishing open standards for agent-to-agent communication, identity, and runtime port...
Hosted agent platforms like Lindy, Relevance AI, and Microsoft Power Virtual Agents handle inbound support tickets, route conversations, and resolve common requests end-to-end.
Research and Knowledge Work
LangChain and LlamaIndex agents combine retrieval over enterprise documents with reasoning to answer complex questions and synthesize multi-source briefings.
Sales and Outbound Workflows
Agents built on Lindy, Relevance AI, and Composio prospect leads, draft outreach, log CRM activity, and orchestrate multi-step sales sequences.
Software Engineering Agents
AutoGPT, AutoGen, and CrewAI-based engineering agents read repositories, draft pull requests, run tests, and iterate on code under human review.
Voice and Telephony Agents
LiveKit and similar runtimes power real-time voice agents that handle phone calls, meetings, and live conversations using streaming LLMs and tool calling.
Internal Operations and IT Automation
Agentic platforms like Restack, Trigger.dev, and Stackmint orchestrate long-running internal workflows across SaaS, databases, and identity systems.
Multi-Agent Team Simulations
CrewAI and AutoGen model role-based teams (researcher, writer, reviewer) collaborating on a single deliverable, producing higher-quality output than a single agent.
Pre-built integrations with other platforms and tools.
LangChain
The most widely used open-source framework for building LLM applications, including chains, agents, retrieval, memory, and tool calling across dozens of providers.
LangGraph
A graph-based runtime from LangChain for stateful, multi-actor, long-running agent workflows with checkpointing and human-in-the-loop.
CrewAI
Multi-agent framework for orchestrating role-based agent crews that collaborate on complex tasks.
AutoGen
Microsoft's open-source framework for building multi-agent conversations with planner, executor, and critic roles.
Composio
Agent execution platform exposing 1000+ apps as governed tools for any agent framework or model.
Letta
Stateful agent platform (formerly MemGPT) focused on long-term memory and persistent agent identity.
OpenAI Assistants
OpenAI's hosted agent runtime with built-in tool use, retrieval, code interpreter, and threads.
Amazon Bedrock Agents
AWS's managed agent runtime over foundation models with action groups, knowledge bases, and guardrails.
aid: agents
name: Agents
description: An index and topic collection covering AI agents, agent frameworks, and agent runtimes that enable autonomous
reasoning, tool use, and multi-step task execution. This profile catalogs the platforms and open-source projects that let
developers build, orchestrate, and deploy LLM-powered agents, including foundational frameworks like LangChain, LangGraph,
CrewAI, AutoGen, and AutoGPT, hosted agent platforms like Lindy, Relevance AI, and Composio, and provider-native agent runtimes
from OpenAI, Anthropic, Google, and Microsoft. Distinct from foundation models and from Agent Skills (capability packages),
the Agents topic focuses on the orchestration layer where planning, memory, tool calling, and execution come together.
type: Index
image: https://kinlane-images.s3.amazonaws.com/shared/apis-json/apis-json-logo.jpg
tags:
- AI Agents
- Agent Frameworks
- Agent Orchestration
- Agent Runtimes
- LLM Orchestration
url: https://raw.githubusercontent.com/api-evangelist/agents/refs/heads/main/apis.yml
created: '2026-05-19'
modified: '2026-05-19'
specificationVersion: '0.19'
apis: []
network:
- name: Agentic AI Foundation
url: https://raw.githubusercontent.com/api-evangelist/agentic-ai-foundation/refs/heads/main/apis.yml
- name: AgentGateway
url: https://raw.githubusercontent.com/api-evangelist/agentgateway/refs/heads/main/apis.yml
- name: AgentQL
url: https://raw.githubusercontent.com/api-evangelist/agentql/refs/heads/main/apis.yml
- name: Amazon Bedrock
url: https://raw.githubusercontent.com/api-evangelist/amazon-bedrock/refs/heads/main/apis.yml
- name: Anthropic
url: https://raw.githubusercontent.com/api-evangelist/anthropic/refs/heads/main/apis.yml
- name: AutoGen
url: https://raw.githubusercontent.com/api-evangelist/autogen/refs/heads/main/apis.yml
- name: AutoGPT
url: https://raw.githubusercontent.com/api-evangelist/autogpt/refs/heads/main/apis.yml
- name: Bifrost
url: https://raw.githubusercontent.com/api-evangelist/bifrost/refs/heads/main/apis.yml
- name: Composio
url: https://raw.githubusercontent.com/api-evangelist/composio/refs/heads/main/apis.yml
- name: CrewAI
url: https://raw.githubusercontent.com/api-evangelist/crewai/refs/heads/main/apis.yml
- name: Google ADK
url: https://raw.githubusercontent.com/api-evangelist/google-adk/refs/heads/main/apis.yml
- name: Google Vertex AI
url: https://raw.githubusercontent.com/api-evangelist/google-vertex-ai/refs/heads/main/apis.yml
- name: Haystack
url: https://raw.githubusercontent.com/api-evangelist/haystack/refs/heads/main/apis.yml
- name: Jentic
url: https://raw.githubusercontent.com/api-evangelist/jentic/refs/heads/main/apis.yml
- name: kagent
url: https://raw.githubusercontent.com/api-evangelist/kagent/refs/heads/main/apis.yml
- name: LangChain
url: https://raw.githubusercontent.com/api-evangelist/langchain/refs/heads/main/apis.yml
- name: LangGraph
url: https://raw.githubusercontent.com/api-evangelist/langgraph/refs/heads/main/apis.yml
- name: Letta
url: https://raw.githubusercontent.com/api-evangelist/letta/refs/heads/main/apis.yml
- name: Lindy
url: https://raw.githubusercontent.com/api-evangelist/lindy/refs/heads/main/apis.yml
- name: LiveKit
url: https://raw.githubusercontent.com/api-evangelist/livekit/refs/heads/main/apis.yml
- name: LlamaIndex
url: https://raw.githubusercontent.com/api-evangelist/llamaindex/refs/heads/main/apis.yml
- name: Microsoft Azure AI Foundry
url: https://raw.githubusercontent.com/api-evangelist/azure-ai-foundry/refs/heads/main/apis.yml
- name: Microsoft Power Virtual Agents
url: https://raw.githubusercontent.com/api-evangelist/microsoft-power-virtual-agents/refs/heads/main/apis.yml
- name: OpenAI
url: https://raw.githubusercontent.com/api-evangelist/openai/refs/heads/main/apis.yml
- name: Portkey
url: https://raw.githubusercontent.com/api-evangelist/portkey/refs/heads/main/apis.yml
- name: Relevance AI
url: https://raw.githubusercontent.com/api-evangelist/relevance-ai/refs/heads/main/apis.yml
- name: Restack
url: https://raw.githubusercontent.com/api-evangelist/restack/refs/heads/main/apis.yml
- name: Stackmint
url: https://raw.githubusercontent.com/api-evangelist/stackmint/refs/heads/main/apis.yml
- name: SuperAgent
url: https://raw.githubusercontent.com/api-evangelist/superagent/refs/heads/main/apis.yml
- name: Toolhouse
url: https://raw.githubusercontent.com/api-evangelist/toolhouse/refs/heads/main/apis.yml
- name: Trigger.dev
url: https://raw.githubusercontent.com/api-evangelist/trigger-dev/refs/heads/main/apis.yml
- name: TrueFoundry
url: https://raw.githubusercontent.com/api-evangelist/truefoundry/refs/heads/main/apis.yml
common:
- type: Portal
url: https://apievangelist.com
- type: GitHubOrganization
url: https://github.com/api-evangelist
- type: Features
data:
- name: Planning and Reasoning Loops
description: Agent frameworks like LangGraph, AutoGen, and CrewAI provide structured loops for planning, reflection, and
multi-step reasoning over user goals, breaking complex tasks into ordered sub-tasks.
- name: Tool Calling and Function Use
description: Agents invoke external functions, APIs, and aggregated tool catalogs through provider-native tool calling,
expanding their reach beyond text generation into real-world action.
- name: Memory and State Management
description: Frameworks like Letta and LangGraph provide persistent memory, conversation state, and long-running execution
context so agents can resume work, remember users, and accumulate knowledge over time.
- name: Multi-Agent Orchestration
description: CrewAI, AutoGen, and LangGraph enable multiple specialized agents to collaborate, delegate sub-tasks, and
reach consensus, modeling teams of role-based workers around a shared goal.
- name: Provider-Native Agent Runtimes
description: OpenAI Assistants, Anthropic Claude tool use, Google ADK, Amazon Bedrock Agents, and Microsoft Azure AI Foundry
expose first-party agent runtimes that handle planning, tool invocation, and state on the model provider side.
- name: Browser and Web Agents
description: Agents like AgentQL and browser-driving frameworks let LLMs navigate real web pages, fill forms, and extract
structured data, extending agent reach to systems that lack APIs.
- name: Observability and Evaluation
description: Platforms like Portkey, TrueFoundry, and Bifrost provide tracing, evaluation, cost tracking, and gateway
routing across agent runs to make agent behavior observable and debuggable.
- name: Open Governance and Interoperability
description: The Linux Foundation Agentic AI Foundation and projects like kagent and AgentGateway are establishing open
standards for agent-to-agent communication, identity, and runtime portability.
- type: UseCases
data:
- name: Customer Support Automation
description: Hosted agent platforms like Lindy, Relevance AI, and Microsoft Power Virtual Agents handle inbound support
tickets, route conversations, and resolve common requests end-to-end.
- name: Research and Knowledge Work
description: LangChain and LlamaIndex agents combine retrieval over enterprise documents with reasoning to answer complex
questions and synthesize multi-source briefings.
- name: Sales and Outbound Workflows
description: Agents built on Lindy, Relevance AI, and Composio prospect leads, draft outreach, log CRM activity, and orchestrate
multi-step sales sequences.
- name: Software Engineering Agents
description: AutoGPT, AutoGen, and CrewAI-based engineering agents read repositories, draft pull requests, run tests,
and iterate on code under human review.
- name: Voice and Telephony Agents
description: LiveKit and similar runtimes power real-time voice agents that handle phone calls, meetings, and live conversations
using streaming LLMs and tool calling.
- name: Internal Operations and IT Automation
description: Agentic platforms like Restack, Trigger.dev, and Stackmint orchestrate long-running internal workflows across
SaaS, databases, and identity systems.
- name: Multi-Agent Team Simulations
description: CrewAI and AutoGen model role-based teams (researcher, writer, reviewer) collaborating on a single deliverable,
producing higher-quality output than a single agent.
- type: Integrations
data:
- name: LangChain
description: The most widely used open-source framework for building LLM applications, including chains, agents, retrieval,
memory, and tool calling across dozens of providers.
- name: LangGraph
description: A graph-based runtime from LangChain for stateful, multi-actor, long-running agent workflows with checkpointing
and human-in-the-loop.
- name: CrewAI
description: Multi-agent framework for orchestrating role-based agent crews that collaborate on complex tasks.
- name: AutoGen
description: Microsoft's open-source framework for building multi-agent conversations with planner, executor, and critic
roles.
- name: Composio
description: Agent execution platform exposing 1000+ apps as governed tools for any agent framework or model.
- name: Letta
description: Stateful agent platform (formerly MemGPT) focused on long-term memory and persistent agent identity.
- name: OpenAI Assistants
description: OpenAI's hosted agent runtime with built-in tool use, retrieval, code interpreter, and threads.
- name: Amazon Bedrock Agents
description: AWS's managed agent runtime over foundation models with action groups, knowledge bases, and guardrails.
maintainers:
- FN: Kin Lane
email: kin@apievangelist.com
humanURL: https://apievangelist.com
Discovery needs no key. Ratings and market analysis are Pro.
Get an API key
Free tier, no form to fill in. Signing in shares your email address with us — we
store it to create your key and to recognise you if you sign in with another
provider. See our Privacy Policy and
Terms.
Kin Score
How this is scored →