Add sample: Unstructured Transform MCP (Strands + AgentCore Gateway)

[dizon-u10d]

Adds a sample under python/03-integrate/protocols/unstructured-transform-mcp/ showing a Strands agent connecting to the hosted Unstructured Transform MCP server over streamable-http (bearer API-key auth) to run an asynchronous partition → enrich → chunk → embed pipeline through MCP tools.

Includes:

• main.py (MCPClient + streamablehttp_client), README.md, pyproject.toml, uv.lock, .env.example
• AGENTCORE_GATEWAY.md — federating Transform MCP through Amazon Bedrock AgentCore Gateway
• A capability comparison with the AWS Labs Document Loader MCP server

Closes #286

Docs: https://docs.unstructured.io/transform/overview · Get started: https://transform.unstructured.io/get-started

🤖 Generated with Claude Code

[github-actions]

Latest scan for commit: f3c2eaf | Updated: 2026-07-13 14:33:58 UTC

✅ Security Scan Report (PR Files Only)

Scanned Files

• python/03-integrate/protocols/unstructured-transform-mcp/.env.example
• python/03-integrate/protocols/unstructured-transform-mcp/AGENTCORE_GATEWAY.md
• python/03-integrate/protocols/unstructured-transform-mcp/README.md
• python/03-integrate/protocols/unstructured-transform-mcp/main.py
• python/03-integrate/protocols/unstructured-transform-mcp/pyproject.toml
• python/03-integrate/protocols/unstructured-transform-mcp/uv.lock

---

Security Scan Results

Critical	High	Medium	Low	Info
0	0	0	0	0

Threshold: High

No security issues detected in your changes. Great job!

This scan only covers files changed in this PR.

[dizon-u10d]

Hi @mvangara10, following up on this sample (Strands agent + Transform MCP over streamable-http, plus an AgentCore Gateway guide).

Ran the connection portion of main.py against the live server: with an invalid token it correctly raises MCPClientInitializationError from a genuine 401 Unauthorized (httpx.HTTPStatusError), confirming the transport/auth wiring is correct end-to-end. Tracking issue is #286 per CONTRIBUTING.md. Happy to adjust anything that doesn't fit sample conventions.

[mvangara10]

@manoj-selvakumar5

[dizon-u10d]

@manoj-selvakumar5 following up if there's anything I should adjust.

[dizon-u10d]

Updated for the recent Transform MCP tool rename: transform_files→start_transform_job, check_transform_status→check_job_status, get_transform_results→get_job_results. Ready for another look whenever you get a chance.

[dizon-u10d]

Small follow-up: corrected the tool count in main.py's docstring (server exposes 7 tools total now, this sample uses the 4 core pipeline ones).

[dizon-u10d]

Another round of tool-surface changes shipped upstream this week (file manager tool added), so I went further than a rename this time: rewrote the docs, notebook, and system/agent prompts to describe the pipeline by behavior (submit → poll → fetch) rather than hardcoding exact tool names, and lean on the toolset-discovery pattern already shown here to find the real names at connect time. Should hold up through future renames/additions without needing another update like this.