A curated, course-oriented index of Amazon Bedrock AgentCore resources: AWS documentation, AWS blog posts, re:Invent videos/transcripts, and relevant GitHub repositories.
Status: local draft prepared from the course corpus. This is ready to publish as a GitHub repo after approval.
Freshness note: the initial index was generated from a 2026-06-14 course corpus. Later AgentCore release-note updates are tracked explicitly in AWS documentation map and later blog ingests are appended to AWS blog posts.
Amazon Bedrock AgentCore is AWS infrastructure for taking AI agents from prototypes to production. The AWS docs describe it as a way to build autonomous agents that reason, use tools, and maintain context for use cases such as customer support, workflow automation, data analysis, and coding assistance. Agents can run serverlessly with isolated sessions, persistent memory, and built-in observability.
AgentCore is not a single agent framework. It is a set of managed production building blocks around agent applications:
- Runtime — host agent code or tools, including HTTP, MCP, A2A, and AG-UI style integrations; supports isolated sessions and long-running/asynchronous work.
- Gateway — expose APIs, Lambda functions, existing services, and MCP servers as governed tools for agents, with authentication and policy controls.
- Identity — centralized identity and credential management for agents, including AWS SigV4, OAuth 2.0, API keys, workload identities, and on-behalf-of token exchange patterns.
- Memory — managed short-term and long-term memory so agents can preserve user/session context and retrieve durable insights over time.
- Browser — secure, isolated browser automation with observability features such as live view, session recording/replay, and CloudWatch metrics.
- Code Interpreter — secure sandboxed code execution for analysis, computation, and agent-generated code workflows.
- Observability — service-provided metrics plus custom traces/spans/logs via CloudWatch and OpenTelemetry-compatible instrumentation.
- Evaluations / Policy — evaluate agent quality and enforce deterministic policy boundaries around tool use and gateway traffic.
Sources: local AgentCore docs recall over the AWS Developer Guide, especially:
- https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/what-is-bedrock-agentcore.html
- https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/runtime-sessions.html
- https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/gateway.html
- https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/identity-overview.html
- https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/memory.html
- https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/browser-tool.html
- https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/code-interpreter-tool.html
- https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/observability.html
- https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/policy.html
- Featured knowledge skill: AWS Bedrock AgentCore Skill
- GitHub repositories
- AWS blog posts
- YouTube / re:Invent videos
- AWS documentation map
- Start with the AgentCore overview and Runtime docs.
- Watch the re:Invent intro and architecture videos: AIM2204 and AIM431.
- Learn the core components: Runtime, Gateway, Identity, Memory, Browser, Code Interpreter, Observability, Evaluations, Policy.
- Use the
awslabs/agentcore-samples, SDK, and CLI repositories for labs. - Pick scenario-specific AWS blog samples for applied modules: SRE, media, finance, healthcare, payments, SaaS, voice, browser automation, or MCP tooling.