Skip to content

Add AWS AgentCore integration blog post#1377

Merged
benfrank241 merged 5 commits intomainfrom
feat/agentcore-blog
May 1, 2026
Merged

Add AWS AgentCore integration blog post#1377
benfrank241 merged 5 commits intomainfrom
feat/agentcore-blog

Conversation

@benfrank241
Copy link
Copy Markdown
Contributor

Summary

Blog post covering persistent memory for AWS AgentCore (Bedrock Agents) with Hindsight integration.

Covers the HindsightRuntimeAdapter architecture, multi-tenant memory scoping with TurnContext, three real-world use cases (customer support, data analysis, code review), and best practices for production deployments.

Test plan

  • Blog builds without errors (npm run build)
  • Images render correctly at /img/blog/agentcore-persistent-memory.png
  • Internal links work correctly
  • Code examples are readable and properly formatted

🤖 Generated with Claude Code

benfrank241 and others added 5 commits May 1, 2026 11:35
Blog covers persistent memory for multi-turn agent workflows, including:
- HindsightRuntimeAdapter architecture (before_turn, run_turn, after_turn)
- TurnContext-based memory scoping for multi-tenant deployments
- Three real-world use cases (customer support, data analysis, code review)
- Configuration examples and best practices

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Corrected all code examples to use actual HindsightRuntimeAdapter API:
- Changed from passing hindsight_client to using configure()
- Fixed TurnContext fields: runtime_session_id, user_id, agent_name, tenant_id, request_id
- Updated run_turn signature: context, payload, agent_callable
- Fixed agent_callable signature: async (payload, memory_context) -> dict
- Updated all three use cases with correct async/await patterns
- Clarified memory scoping with actual bank ID generation
- Verified all examples against actual API (tests pass)

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Major revisions based on feedback:

1. Added 'Hindsight vs AgentCore Memory' comparison section explaining:
   - Why Hindsight over AgentCoreMemory (higher-level ops, richer retrieval, automation)
   - When to use each service (choice matrix with 5 differentiators)
   - Automation differences (Hindsight reads+writes, AgentCore write-only)

2. Simplified tone from infrastructure-focused to developer-focused:
   - Rewrote 'Why Your Agents Need Memory' with customer context example
   - Updated 'How Memory Works' to emphasize developer experience
   - Renamed sections for clarity ('Adding Memory to Your Agent' vs 'Setting Up')
   - Practical best practices for developers building agents
   - Developer-centric troubleshooting guidance

3. Added 'Where to Deploy Hindsight' section:
   - Hindsight Cloud (managed) vs Self-Hosted OSS option
   - Clear recommendation: start with Cloud, migrate if needed

All code examples verified and unchanged. Blog now targets AWS developers building agents, not infrastructure engineers.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Move cover image to start of content (after truncate marker) to match other blog posts.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Change publication date from April 29 to May 1, 2026 to align with current review timeline.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
@benfrank241 benfrank241 merged commit 4986e8e into main May 1, 2026
54 of 55 checks passed
@benfrank241 benfrank241 deleted the feat/agentcore-blog branch May 1, 2026 19:06
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant