docs: add agentic patterns overview to cookbook

[nuthalapativarun]

Issue Link (REQUIRED)

Fixes #328

Summary

Changes

Adds docs/src/content/docs/cookbook/patterns/agentic-patterns.mdx — a new cookbook page covering the four core agentic patterns for Agent Squad. Updates astro.config.mjs to surface the page in the sidebar under a new Agentic Patterns heading.

The four patterns documented:

1. Multi-Agent Routing — classifier picks the right specialist agent
2. Supervisor Pattern — lead agent coordinates a parallel team
3. Sequential Chain Pattern — ChainAgent pipelines output through stages
4. RAG Pattern — agent grounded in a retriever/knowledge base

Each pattern includes side-by-side TypeScript and Python examples and a brief "when to use" note. The page closes with a note on combining patterns.

User experience

Before: no single place explaining the architectural patterns available in Agent Squad.

After: users land on Agentic Patterns Overview from the sidebar and can immediately see code examples for their chosen pattern in their preferred language.

Checklist

• I have performed a self-review of this change
• Changes have been tested
• Changes are documented
• I have linked this PR to an existing issue (required)

Acknowledgment

By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.

[cornelcroi]

Hey @nuthalapativarun, the structure and descriptions are good but several code examples are broken and will fail if someone copies them.

The TypeScript RAG example calls AmazonKnowledgeBasesRetriever with just an options object but the actual constructor requires a BedrockAgentRuntimeClient as the first argument, also the field name is retrievalConfiguration not retrievalConfig. The Python routing example does response, _ = await orchestrator.route_request(...) but route_request returns a single AgentResponse not a tuple, so that will throw a ValueError. The Python RAG example imports AmazonKnowledgeBasesRetrieverConfig which does not exist, the correct class is AmazonKnowledgeBasesRetrieverOptions, and the options kwarg is wrong too — it should be retrievalConfiguration not retrieval_config.

Also worth fixing in the Supervisor examples for both languages — name and description passed to SupervisorAgentOptions are silently overwritten by the lead agent values, so either remove those fields from the example or add a comment explaining it.