Skip to content
Kitbay
SkillOfficial

amazon-bedrock

Apache-2.0Updated on Nov 11, 20263,464GitHub starsIncludes scripts

A core skill from AWS's agent toolkit for generative AI work on Amazon Bedrock.

InstallRepository Checked on Nov 13, 2026

What the skill teaches

A core skill from AWS's agent toolkit for generative AI work on Amazon Bedrock. It covers calling models through the Converse API or InvokeModel, retrieval and RAG with Knowledge Bases including chunking choices, Bedrock Agents, Guardrails, and AgentCore, including the managed Harness agent loop and moving Bedrock Agents onto it. It also helps choose between models such as Claude, Llama, Nova and Titan, move between Claude generations, use prompt caching, diagnose throttling, quota and access-denied errors, track cost, and set up AgentCore Payments. It comes with reference files and a Python script that snapshots an existing Bedrock agent for migration.

Capabilities

  • Invoke Bedrock models with Converse or InvokeModel
  • Build RAG with Bedrock Knowledge Bases
  • Configure Bedrock Guardrails
  • Migrate a Bedrock Agent to AgentCore
  • Diagnose Bedrock throttling and access errors

Other MCP servers and skills for Amazon Web Services