package
SDK for building AI agents with Claude Code's capabilities. Programmatically interact with Claude to build autonomous agents that can understand codebases, edit files, and execute workflows.
npm install @anthropic-ai/claude-agent-sdkGitHub-flavored package documentation.
Across all versions.
Download trend unavailable.
Runtime metadata.
Tagged topics.
Estimated unpacked size across 1 packages.
5.1 MB
Includes dependencies.
dylanc-anthropic
dylanc@anthropic.com
ejlangev-ant
ethan.langevin@anthropic.com
jv-anthropic
jv@anthropic.com
ollie-ant-2025
ollie@anthropic.com
packy-anthropic
packy@anthropic.com
claude-code-npm
claude-code-npm@anthropic.com
noahz-anthropic
noahz@anthropic.com
sbidasaria
sbidasaria@anthropic.com
wolffiex
wolffiex@anthropic.com
meorphis-anthropic
meorphis@anthropic.com
felixrieseberg-anthropic
f@anthropic.com
joan-anthropic
joan@anthropic.com
The Claude Agent SDK enables you to programmatically build AI agents with Claude Code's capabilities. Create autonomous agents that can understand codebases, edit files, run commands, and execute complex workflows.
Learn more in the official documentation.
Install the Claude Agent SDK:
npm install @anthropic-ai/claude-agent-sdk/core entry (for apps that bundle the SDK)@anthropic-ai/claude-agent-sdk/core exports the same query(), startup(), prewarm(), tool(), createSdkMcpServer(), renameSession()/tagSession()/archiveSession()/unarchiveSession()/deleteSession(), resolveSettings()/filterEscalatingDefaultMode()/matchDeniedModels() and runtime constants as the root entry, and nothing else — no session listing/reading/import helpers — so its eagerly loaded code is about a quarter of the root's; the settings resolver behind resolveSettings(), the matcher behind matchDeniedModels() and the code behind the session mutations load on first call. Unlike the root entry it does not inline the SDK's peer dependencies: zod (^4, as declared in this package's peerDependencies) and @modelcontextprotocol/sdk (^1.29) are imported from your install — specifically the zod/v4 subpath, plus zod/v3 when the settings resolver loads — which is what you want when your bundler already carries them. (zod 3.25+, which carries the v4 engine under zod/v4, also works in practice, but ^4 is the supported range.) Import from either the root or /core in a given process, not both — they are separate bundles, and mixing them loads two copies of the SDK's classes and state. Its types require TypeScript 5.0+.
bun build --compile)When bundling your application into a single executable with bun build --compile, the SDK cannot resolve the native CLI binary at runtime — require.resolve doesn't work from inside the compiled $bunfs (or B:\~BUN\... on Windows) virtual filesystem.
Embed the platform-specific binary as a file asset, extract it to a real path, and pass it explicitly:
import binPath from '@anthropic-ai/claude-agent-sdk-darwin-arm64/claude' with { type: 'file' }
import { extractFromBunfs } from '@anthropic-ai/claude-agent-sdk/extract'
import { query } from '@anthropic-ai/claude-agent-sdk'
const cliPath = extractFromBunfs(binPath)
for await (const message of query({
prompt: '…',
options: { pathToClaudeCodeExecutable: cliPath },
})) { /* … */ }Each compiled executable embeds one platform's binary, matching your --target. Cross-compiling requires installing the non-matching platform package (e.g. npm install @anthropic-ai/claude-agent-sdk-linux-x64 --force). On Windows the binary subpath is /claude.exe (e.g. @anthropic-ai/claude-agent-sdk-win32-x64/claude.exe).
The Claude Code SDK is now the Claude Agent SDK. Please check out the migration guide for details on breaking changes.
We welcome your feedback. File a GitHub issue to report bugs or request features.
Join the Claude Developers Discord to connect with other developers building with the Claude Agent SDK. Get help, share feedback, and discuss your projects with the community.
When you use the Claude Agent SDK, we collect feedback, which includes usage data (such as code acceptance or rejections), associated conversation data, and user feedback submitted via the /bug command.
See our data usage policies.
We have implemented several safeguards to protect your data, including limited retention periods for sensitive information and restricted access to user session data.
For full details, please review our Commercial Terms of Service and Privacy Policy.