Agent SDK
An SDK provided by Anthropic that allows developers to invoke the Claude Code runtime programmatically. It supports tool-call loops, context persistence, and multi-step reasoning coordination, enabling the embedding of Claude Agent capabilities into custom systems and automated workflows.
Core Facts
OpenAI's Agent SDK takes a more engineering-oriented approach, natively supporting agent-to-agent hand-off mechanisms and built-in safety guardrails.
Agent SDK is a development toolkit written in TypeScript and Python, and is the core component powering the Claude Code backend
Timeline (last 90 days)
In multi-agent architectures, the choice between serial and parallel invocation simultaneously affects both cost and latency metrics.
The CCAR-F exam tests decision-making judgment when building applications using Claude Code, Agent SDK, API, and MCP.
Spotify's backend coding agent is built on Agent SDK and adopts a subagent (multi-agent) pattern at the underlying level
Spotify's backend coding agent built on Agent SDK automatically merges more than 650 Pull Requests per month
Anthropic's Agent SDK has integration support with observability platforms such as LangSmith and Weave
Running claude -p, Agent SDK sessions, and cloud sessions will load project-scope servers directly without prompting, bypassing the approval prompt
Anthropic's internal CoWork product is built on the Agent SDK.
Through the Agent SDK, Claude Code's runtime can be called directly without relying on the official CLI or pre-built tools.
Connecting to the Agent SDK requires only an API Key.
OpenAI's Agent SDK takes a more engineering-oriented approach, natively supporting agent-to-agent hand-off mechanisms and built-in safety guardrails.
3 more timeline events
All Facts (18)
OpenAI's Agent SDK takes a more engineering-oriented approach, natively supporting agent-to-agent hand-off mechanisms and built-in safety guardrails.
65%VerifiedAgent SDK is a development toolkit written in TypeScript and Python, and is the core component powering the Claude Code backend
65%UnverifiedAnthropic的Agent SDK支持定义智能体的行动空间和终止条件，内置安全约束和行为边界
85%UnverifiedOpenHands的核心架构基于Agent SDK引擎，利用ReAct框架和Function Calling机制实现代理推理和执行
85%Unverified智能体任务的Token消耗量通常是普通对话的数十倍
75%UnverifiedThird-party applications such as Conductor and OpenCloud are developed based on the Claude Agent SDK
60%UnverifiedIn multi-agent architectures, the choice between serial and parallel invocation simultaneously affects both cost and latency metrics.
50%UnverifiedThe CCAR-F exam tests decision-making judgment when building applications using Claude Code, Agent SDK, API, and MCP.
50%UnverifiedSpotify's backend coding agent built on Agent SDK automatically merges more than 650 Pull Requests per month
50%UnverifiedSpotify's backend coding agent is built on Agent SDK and adopts a subagent (multi-agent) pattern at the underlying level
50%UnverifiedAnthropic's Agent SDK has integration support with observability platforms such as LangSmith and Weave
50%UnverifiedRunning claude -p, Agent SDK sessions, and cloud sessions will load project-scope servers directly without prompting, bypassing the approval prompt
50%UnverifiedAnthropic's internal CoWork product is built on the Agent SDK.
50%UnverifiedThrough the Agent SDK, Claude Code's runtime can be called directly without relying on the official CLI or pre-built tools.
50%UnverifiedConnecting to the Agent SDK requires only an API Key.
50%UnverifiedMainstream frameworks such as LangChain, AutoGen, and Google's Agent SDK constrain agent risks through mechanisms such as tool whitelists and pre-execution human confirmation.
50%UnverifiedThe update fixes an issue where claude -p and Agent SDK sessions could hang without output after an internal error; they now report the error and exit with exit code 1.
50%UnverifiedAs pay-per-use models become increasingly mainstream, inefficient token usage will be directly reflected in bills.
50%