ydc-openai-agent-sdk-integration_skill

This skill assists developers integrate OpenAI Agents SDK with You.com MCP, enabling hosted or streamable HTTP MCP setups for Python and TypeScript.

TypeScript

4

GitHub Stars

1

Bundled Files

2 months ago

Catalog Refreshed

4 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill youdotcom-oss/agent-skills --skill ydc-openai-agent-sdk-integration

SKILL.md17.5 KB

Overview

This skill integrates the OpenAI Agents SDK with You.com MCP so agents can call You.com tools from Python or TypeScript. It provides ready-to-run templates for both Hosted MCP and Streamable HTTP modes, plus validation checks and environment guidance. Use it to wire @openai/agents or openai-agents into your MCP-enabled agent workflows quickly.

How this skill works

The integration configures an Agent instance with either a HostedMCPTool (OpenAI-managed) or an MCPServerStreamableHttp (self-managed) and injects You.com authorization headers using YDC_API_KEY. It validates required environment variables (YDC_API_KEY and OPENAI_API_KEY), installs the proper package (openai-agents or @openai/agents), and offers complete Python and TypeScript templates to run immediately. After setup, the agent can discover You.com tools like web search and content extraction.

When to use it

You mention OpenAI Agents SDK or OpenAI agents and need You.com tool access
Building production apps that prefer Hosted MCP for simplicity and reliability
Implementing custom transport, timeouts, or retries with Streamable HTTP
Adding You.com tool discovery (mcp__ydc__you_search, mcp__ydc__you_contents) to agent workflows
Integrating @openai/agents or openai-agents into developer tools or CI pipelines

Best practices

Choose Hosted MCP for production to offload connection management and reduce boilerplate
Use Streamable HTTP when you need custom headers, timeouts, or to run MCP in your infra
Store keys in environment variables (.env) and validate YDC_API_KEY and OPENAI_API_KEY at startup
Add require_approval or tool-level restrictions if agents will run untrusted prompts
Follow the validation checklist: package installed, env vars set, templates integrated, and Authorization headers configured

Example use cases

Python agent that searches the web for latest AI news using Hosted MCP template
TypeScript dev tool that extracts web page content via mcp__ydc__you_contents using Streamable HTTP
CI job that runs agent tests against You.com tools with mocked MCP server
Embedding You.com search and extraction into an AI development platform using @openai/agents

FAQ

Install openai-agents for Python or @openai/agents for TypeScript depending on your language choice.

When should I use Hosted MCP vs Streamable HTTP?

Use Hosted MCP for simplicity, reliability, and lower maintenance; use Streamable HTTP when you need custom transport, headers, or local control.

What environment variables are required?

Set YDC_API_KEY (You.com) and OPENAI_API_KEY (OpenAI). Validate both before running the agent.