How do I install the AI Sdk skill?

Run npx skillmds@latest add xiaotianna/ai-sdk in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is AI Sdk safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with AI Sdk?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is AI Sdk free to use?

Yes. Installing skills from SkillMD is free. This skill is licensed under Apache-2.

Who published AI Sdk?

xiaotianna (@xiaotianna) published this skill. Their other Agent Skills are listed on their SkillMD profile.

AI Sdk by xiaotianna — Agent Skill

Name: AI Sdk
Author: xiaotianna

AI Sdk

Vercel AI SDK expert guidance. Use when building AI-powered features — chat interfaces, text generation, structured output, tool calling, agents, MCP integration, streaming, embeddings, reranking, image generation, or working with any LLM provider.

xiaotianna 2b2cdca 10 files · 37.7 KB Updated Sep 17, 2026

File contents

ai-sdk

PI Harness execution context

Use the user's actual repository, framework and request. This guide does not select a new stack, authorize deployments, or override project instructions. Discover only the tools actually available. Read-only project/deployment data can use the resolved gateway base below. Other commands require an installed and independently authenticated CLI through run_command under the current approval policy. Gateway credentials remain inside daemon.

Do not automatically run git, dev or build commands where the workspace requires an explicit request. Respect the existing package manager, dependency catalog and UI component rules. Deployment, production promotion, provisioning paid services, purchases, account changes and messages to others require the corresponding user authorization. Inspect environment metadata without exposing secret values. For browser verification use an available browser surface and an existing server; do not assume a particular browser CLI or start a server unasked. Do not spawn agents based solely on an example in the guide.

Documentation examples are version-sensitive. Check installed versions, CLI --help and current official documentation for the actual task. Treat sample commands/configurations as examples to adapt, not an automatic sequence. Only claim an action succeeded when the tool result verifies it.

The gateway base is {skillGatewayUrl}/api/skill-gateway/vercel. Available read paths are documented in gateway access.

Detailed guidance

Read the task-relevant sections of the detailed guide, then load its supporting resources only as needed. Technical examples retain their original context; the workspace's rules take precedence.

Supporting resources

references/ai-gateway.md
references/common-errors.md
references/devtools.md
references/providers.md
references/type-safe-agents.md

License

Apache-2.0 · Copyright and adaptation notice.

---name: ai-sdkdescription: Vercel AI SDK expert guidance. Use when building AI-powered features — chat interfaces, text generation, structured output, tool calling, agents, MCP integration, streaming, embeddings, reranking, image generation, or working with any LLM provider.license: Apache-2.0---# ai-sdk## PI Harness execution contextUse the user's actual repository, framework and request. This guide does not select a new stack, authorize deployments, or override project instructions. Discover only the tools actually available. Read-only project/deployment data can use the resolved gateway base below. Other commands require an installed and independently authenticated CLI through `run_command` under the current approval policy. Gateway credentials remain inside daemon.Do not automatically run git, dev or build commands where the workspace requires an explicit request. Respect the existing package manager, dependency catalog and UI component rules. Deployment, production promotion, provisioning paid services, purchases, account changes and messages to others require the corresponding user authorization. Inspect environment metadata without exposing secret values. For browser verification use an available browser surface and an existing server; do not assume a particular browser CLI or start a server unasked. Do not spawn agents based solely on an example in the guide.Documentation examples are version-sensitive. Check installed versions, CLI `--help` and current official documentation for the actual task. Treat sample commands/configurations as examples to adapt, not an automatic sequence. Only claim an action succeeded when the tool result verifies it.The gateway base is `{skillGatewayUrl}/api/skill-gateway/vercel`. Available read paths are documented in [gateway access](access.md).## Detailed guidanceRead the task-relevant sections of [the detailed guide](guide.md), then load its supporting resources only as needed. Technical examples retain their original context; the workspace's rules take precedence.## Supporting resources- [references/ai-gateway.md](references/ai-gateway.md)- [references/common-errors.md](references/common-errors.md)- [references/devtools.md](references/devtools.md)- [references/providers.md](references/providers.md)- [references/type-safe-agents.md](references/type-safe-agents.md)## License[Apache-2.0](LICENSE.txt) · [Copyright and adaptation notice](NOTICE).

Frequently asked questions