# MCP: Build Rich-Context AI Apps with Anthropic **URL:** https://community.deeplearning.ai/c/short-course-q-a/mcp-build-rich-context-ai-apps-with-anthropic/525.md [Latest](https://community.deeplearning.ai/latest.md) · [Categories](https://community.deeplearning.ai/categories.md) · [Tags](https://community.deeplearning.ai/tags.md) --- ## [Update in notebooks L4, L6, L7 - due to mcp changes in late July 2026](https://community.deeplearning.ai/t/update-in-notebooks-l4-l6-l7-due-to-mcp-changes-in-late-july-2026/894597)
**Author:** [@lesly.zerna](https://community.deeplearning.ai/u/lesly.zerna)\ **Replies:** 0\ **Last updated:** [August 13, 2026, 5:23pm UTC](https://community.deeplearning.ai/t/update-in-notebooks-l4-l6-l7-due-to-mcp-changes-in-late-july-2026/894597 "2026-08-13T17:23:57Z")
Quick status: the errors such as mcp.server.fastmcp import errors, McpError import errors, tool\_use/tool\_result mismatches were due to a breaking update released by MCP on 2026-07-28, which changed parts of the SDK the c… --- ## [L 7Code fix to host research\_server.py in Render](https://community.deeplearning.ai/t/l-7code-fix-to-host-research-server-py-in-render/820254)
**Author:** [@fraserstreet](https://community.deeplearning.ai/u/fraserstreet)\ **Replies:** 3\ **Last updated:** [August 13, 2026, 5:17pm UTC](https://community.deeplearning.ai/t/l-7code-fix-to-host-research-server-py-in-render/820254 "2026-08-13T17:17:58Z")
Hi all, In the final exercise where you run the server in the Render cloud service, I found that you need to change the instantiation of the mcp object around line 10 to “mcp = FastMCP(“research”, host=“0.0.0.0”, port=1… --- ## [MCP Server Error](https://community.deeplearning.ai/t/mcp-server-error/894443)
**Author:** [@ecocarlisle](https://community.deeplearning.ai/u/ecocarlisle)\ **Replies:** 4\ **Last updated:** [August 13, 2026, 5:16pm UTC](https://community.deeplearning.ai/t/mcp-server-error/894443 "2026-08-13T17:16:47Z")
I get the following error when I try to connect: Traceback (most recent call last): File “/home/jovyan/work/L4/mcp\_project/research\_server.py”, line 6, in from mcp.server.fastmcp import FastMCP ModuleNotFoundError: … --- ## [L4 - Unable to spin up a MCP Inspector](https://community.deeplearning.ai/t/l4-unable-to-spin-up-a-mcp-inspector/825660)
**Author:** [@Balaji.Illur](https://community.deeplearning.ai/u/Balaji.Illur)\ **Replies:** 6\ **Last updated:** [August 13, 2026, 4:53pm UTC](https://community.deeplearning.ai/t/l4-unable-to-spin-up-a-mcp-inspector/825660 "2026-08-13T16:53:48Z")
Hi Team, I am getting below error in the terminal once I try to connect to MCP server via Model inspector. Please need guidance on the same. @TMosh , request you look at this issue. Looks like is there a potential libr… --- ## [L7 : mcp\_chatbot.py fails with "No module named 'mcp.server.fastmcp'"](https://community.deeplearning.ai/t/l7-mcp-chatbot-py-fails-with-no-module-named-mcp-server-fastmcp/894529)
**Author:** [@aravindagn](https://community.deeplearning.ai/u/aravindagn)\ **Replies:** 5\ **Last updated:** [August 13, 2026, 4:48pm UTC](https://community.deeplearning.ai/t/l7-mcp-chatbot-py-fails-with-no-module-named-mcp-server-fastmcp/894529 "2026-08-13T16:48:22Z")
The problem The course was written for version 1.x of the mcp Python package. The project’s pyproject.toml doesn’t set an upper limit on that version, so uv now installs mcp 2.0, which renamed and moved things. The less… --- ## [L6 - MCP Server Failing Outdated MCP SDK](https://community.deeplearning.ai/t/l6-mcp-server-failing-outdated-mcp-sdk/894445)
**Author:** [@ecocarlisle](https://community.deeplearning.ai/u/ecocarlisle)\ **Replies:** 7\ **Last updated:** [August 13, 2026, 4:47pm UTC](https://community.deeplearning.ai/t/l6-mcp-server-failing-outdated-mcp-sdk/894445 "2026-08-13T16:47:59Z")
The MCP SDK needs to be updated as it is breaking with the current client code. Traceback (most recent call last): File “/home/jovyan/.cache/uv/archive-v0/wUt8Jq4wjzWGjjMfjMAC8/bin/mcp-server-fetch”, line 6, in from m… --- ## [Lesson 3: Chatbot Example notebook is working now!](https://community.deeplearning.ai/t/lesson-3-chatbot-example-notebook-is-working-now/894596)
**Author:** [@lesly.zerna](https://community.deeplearning.ai/u/lesly.zerna)\ **Replies:** 0\ **Last updated:** [August 13, 2026, 4:21pm UTC](https://community.deeplearning.ai/t/lesson-3-chatbot-example-notebook-is-working-now/894596 "2026-08-13T16:21:15Z")
Thanks everyone who reported issues on this lesson. The updates have been pushed to the platform. Note: the def process\_query(query) was refactored. Thanks everyone for their suggestions! Happy Learning! Lesly - … --- ## [The code for chatbot seems problematic, in L3](https://community.deeplearning.ai/t/the-code-for-chatbot-seems-problematic-in-l3/819697)
**Author:** [@yagneshicp](https://community.deeplearning.ai/u/yagneshicp)\ **Replies:** 2\ **Last updated:** [August 13, 2026, 4:18pm UTC](https://community.deeplearning.ai/t/the-code-for-chatbot-seems-problematic-in-l3/819697 "2026-08-13T16:18:23Z")
ChatGPT - Code Issue Analysis - Just see this chat, response, and you will get to know what I am saying --- ## [L3 - Authentication Error](https://community.deeplearning.ai/t/l3-authentication-error/857048)
**Author:** [@diwpuri](https://community.deeplearning.ai/u/diwpuri)\ **Replies:** 2\ **Last updated:** [August 13, 2026, 4:17pm UTC](https://community.deeplearning.ai/t/l3-authentication-error/857048 "2026-08-13T16:17:11Z")
I imported the notebook L3. When I run the cell chat\_loop() I am getting following error "Error: “Could not resolve authentication method. Expected either api\_key or auth\_token to be set. Or for one of the X-Api-Key or A… --- ## [Code fix for Error Code 400 in labs](https://community.deeplearning.ai/t/code-fix-for-error-code-400-in-labs/894175)
**Author:** [@Chris\_Falter](https://community.deeplearning.ai/u/Chris_Falter)\ **Replies:** 2\ **Last updated:** [August 13, 2026, 4:16pm UTC](https://community.deeplearning.ai/t/code-fix-for-error-code-400-in-labs/894175 "2026-08-13T16:16:19Z")
Continuing the discussion from Still facing the error code: 400 in L3: I debugged the code by printing the messages, and discovered that a single tool\_use message was being incorrectly passed multiple times to the model… --- ## [Still facing the error code: 400 in L3](https://community.deeplearning.ai/t/still-facing-the-error-code-400-in-l3/891748)
**Author:** [@Yuxin\_Xu](https://community.deeplearning.ai/u/Yuxin_Xu)\ **Replies:** 10\ **Last updated:** [August 13, 2026, 4:16pm UTC](https://community.deeplearning.ai/t/still-facing-the-error-code-400-in-l3/891748 "2026-08-13T16:16:05Z")
--- ## [Error code: 429 - {'error': {'message': 'exceeded quota for this month'}}](https://community.deeplearning.ai/t/error-code-429-error-message-exceeded-quota-for-this-month/893898)
**Author:** [@Ben\_Majer](https://community.deeplearning.ai/u/Ben_Majer)\ **Replies:** 4\ **Last updated:** [August 10, 2026, 5:07am UTC](https://community.deeplearning.ai/t/error-code-429-error-message-exceeded-quota-for-this-month/893898 "2026-08-10T05:07:42Z")
I’m bumping into an Error code: 429 - {‘error’: {‘message’: ‘exceeded quota for this month’}} that persists from l6 to l7, while we are only midway the month --- ## [Question on MCP Server Orchestration at Scale](https://community.deeplearning.ai/t/question-on-mcp-server-orchestration-at-scale/883343)
**Author:** [@BeyersStrydom](https://community.deeplearning.ai/u/BeyersStrydom)\ **Replies:** 2\ **Last updated:** [July 6, 2026, 9:35pm UTC](https://community.deeplearning.ai/t/question-on-mcp-server-orchestration-at-scale/883343 "2026-07-06T21:35:36Z")
Dear DeepLearning.AI Community, I recently completed the MCP: Build Rich-Content Apps… short course and was really impressed by the developments! I have a question/challenge I’d like to raise to the community: As the n… --- ## [What is the difference between tools and resources in the context of APIs?](https://community.deeplearning.ai/t/what-is-the-difference-between-tools-and-resources-in-the-context-of-apis/819016)
**Author:** [@yagneshicp](https://community.deeplearning.ai/u/yagneshicp)\ **Replies:** 2\ **Last updated:** [April 15, 2026, 7:19am UTC](https://community.deeplearning.ai/t/what-is-the-difference-between-tools-and-resources-in-the-context-of-apis/819016 "2026-04-15T07:19:32Z")
In this part, it is written that API responses are resources. We all know that API calling is part of a tool. So what’s the fundamental difference between tools and resources here? Thanks. --- ## [Still facing the error code: 400 in both L3 and L5](https://community.deeplearning.ai/t/still-facing-the-error-code-400-in-both-l3-and-l5/890003)
**Author:** [@agparth](https://community.deeplearning.ai/u/agparth)\ **Replies:** 4\ **Last updated:** [March 9, 2026, 3:39pm UTC](https://community.deeplearning.ai/t/still-facing-the-error-code-400-in-both-l3-and-l5/890003 "2026-03-09T15:39:32Z")
--- ## [MCP:L5 fixes in mcp\_chatbot.py to run it locally](https://community.deeplearning.ai/t/mcp-l5-fixes-in-mcp-chatbot-py-to-run-it-locally/889719)
**Author:** [@Lauyao](https://community.deeplearning.ai/u/Lauyao)\ **Replies:** 7\ **Last updated:** [March 5, 2026, 1:21am UTC](https://community.deeplearning.ai/t/mcp-l5-fixes-in-mcp-chatbot-py-to-run-it-locally/889719 "2026-03-05T01:21:30Z")
Hello. On this short course MCP: Build Rich-Context AI Apps with Anthropic i have the following error in L5 Creating MCP client. when launching the mcp\_chatbot.py given in L5 I have the following error message when aski… --- ## ['claude-3-7-sonnet-20250219' is deprecated](https://community.deeplearning.ai/t/claude-3-7-sonnet-20250219-is-deprecated/889150)
**Author:** [@GoodVision](https://community.deeplearning.ai/u/GoodVision)\ **Replies:** 2\ **Last updated:** [February 19, 2026, 10:12pm UTC](https://community.deeplearning.ai/t/claude-3-7-sonnet-20250219-is-deprecated/889150 "2026-02-19T22:12:26Z")
Which models should we switch to since current model gives message ‘claude-3-7-sonnet-20250219’ is deprecated and will reach end-of-life on February 19th, 2026 --- ## [You have reached your specified API usage limits](https://community.deeplearning.ai/t/you-have-reached-your-specified-api-usage-limits/880150)
**Author:** [@stainsb](https://community.deeplearning.ai/u/stainsb)\ **Replies:** 11\ **Last updated:** [February 19, 2026, 2:03pm UTC](https://community.deeplearning.ai/t/you-have-reached-your-specified-api-usage-limits/880150 "2026-02-19T14:03:52Z")
In the Chatbot Example section, step #13, I got the following error which happened on the first execution (i.e. I wasn’t running a bunch of examples). Error: Error code: 400 - {‘type’: ‘error’, ‘error’: {‘type’: ‘invali… --- ## [Chatbot code errors out](https://community.deeplearning.ai/t/chatbot-code-errors-out/888009)
**Author:** [@nmurugesh](https://community.deeplearning.ai/u/nmurugesh)\ **Replies:** 2\ **Last updated:** [February 18, 2026, 7:16pm UTC](https://community.deeplearning.ai/t/chatbot-code-errors-out/888009 "2026-02-18T19:16:17Z")
The chatbot code fails because the API call is within the for loop --- ## [Unable to set and get Inspector Proxy Address](https://community.deeplearning.ai/t/unable-to-set-and-get-inspector-proxy-address/819599)
**Author:** [@Neetika\_Jain](https://community.deeplearning.ai/u/Neetika_Jain)\ **Replies:** 2\ **Last updated:** [February 17, 2026, 7:49pm UTC](https://community.deeplearning.ai/t/unable-to-set-and-get-inspector-proxy-address/819599 "2026-02-17T19:49:18Z")
Unable to set and get Inspector Proxy Address while running MCP server and client (“Inspector Proxy Address that you need to specify under configuration in the inspector UI:”) print(os.environ.get(‘DLAI\_LOCAL\_URL’).for… --- ## [Where to find class notes?](https://community.deeplearning.ai/t/where-to-find-class-notes/887300)
**Author:** [@jia.gu12345](https://community.deeplearning.ai/u/jia.gu12345)\ **Replies:** 2\ **Last updated:** [January 17, 2026, 9:56pm UTC](https://community.deeplearning.ai/t/where-to-find-class-notes/887300 "2026-01-17T21:56:38Z")
I realized regular classes and industry certification classes have downloadable lecture notes. This is the 1st short course I’m taking, and I’m unable to find the lecture notes. Does anyone else know where to locate them… --- ## [MCP Tools vs Resources: When does data retrieval become a "tool" operation?](https://community.deeplearning.ai/t/mcp-tools-vs-resources-when-does-data-retrieval-become-a-tool-operation/852579)
**Author:** [@nadba](https://community.deeplearning.ai/u/nadba)\ **Replies:** 1\ **Last updated:** [December 10, 2025, 2:13pm UTC](https://community.deeplearning.ai/t/mcp-tools-vs-resources-when-does-data-retrieval-become-a-tool-operation/852579 "2025-12-10T14:13:37Z")
Referring to MCP architecture lesson at time 2:10. What are the practical differences between MCP Tools and Resources when something like ‘search/retrieve’ feels like passive data access but is classified as a Tool? In… --- ## [How I can download the resources of course?](https://community.deeplearning.ai/t/how-i-can-download-the-resources-of-course/818095)
**Author:** [@Manh\_Truong\_Do](https://community.deeplearning.ai/u/Manh_Truong_Do)\ **Replies:** 4\ **Last updated:** [October 17, 2025, 4:30pm UTC](https://community.deeplearning.ai/t/how-i-can-download-the-resources-of-course/818095 "2025-10-17T16:30:32Z")
I want to download all of them? How can I do? --- ## [Stuck at configuring transport](https://community.deeplearning.ai/t/stuck-at-configuring-transport/853892)
**Author:** [@Margherita\_Di\_Leo](https://community.deeplearning.ai/u/Margherita_Di_Leo)\ **Replies:** 0\ **Last updated:** [July 17, 2025, 4:28pm UTC](https://community.deeplearning.ai/t/stuck-at-configuring-transport/853892 "2025-07-17T16:28:49Z")
MCP is great and I can’t wait to customise my app. But so far I’m stuck at configuring the transport protocol. If I use stdio, I get a few errors like n the screenshot: And I also tried both the sse nd the streamab… --- ## [Await session.call\_tool(tool\_name, arguments=tool\_args) not working in the process query](https://community.deeplearning.ai/t/await-session-call-tool-tool-name-arguments-tool-args-not-working-in-the-process-query/842593)
**Author:** [@bkapmouo](https://community.deeplearning.ai/u/bkapmouo)\ **Replies:** 0\ **Last updated:** [June 30, 2025, 10:55am UTC](https://community.deeplearning.ai/t/await-session-call-tool-tool-name-arguments-tool-args-not-working-in-the-process-query/842593 "2025-06-30T10:55:11Z")
Hi All, I inspired myself from the MCP client code to build a streamlit chatbothat would connect to the following mcp servers. SQLITE BRAVE SEARCH FILE SYSTEM Everything seems to work but whenever I write a prompt in… --- ## [MCP server on RENDER](https://community.deeplearning.ai/t/mcp-server-on-render/842579)
**Author:** [@Marko\_B](https://community.deeplearning.ai/u/Marko_B)\ **Replies:** 0\ **Last updated:** [June 30, 2025, 10:44am UTC](https://community.deeplearning.ai/t/mcp-server-on-render/842579 "2025-06-30T10:44:10Z")
The MCP research\_server.py is running successfully on RENDER. I connect my local mcp\_chatbot.by to the remote server by setting the SERVER URL to the RENDER address in sse\_transport = await self.exit\_stack.enter\_async\_c… --- ## [About how to connect my chatbot to mcp\_server](https://community.deeplearning.ai/t/about-how-to-connect-my-chatbot-to-mcp-server/839690)
**Author:** [@Lambda1415926](https://community.deeplearning.ai/u/Lambda1415926)\ **Replies:** 0\ **Last updated:** [June 25, 2025, 8:39am UTC](https://community.deeplearning.ai/t/about-how-to-connect-my-chatbot-to-mcp-server/839690 "2025-06-25T08:39:55Z")
I’ve open the chat\_bot in L7 and I modified the code( sse\_transport and so on…) the chat\_bot file work just fine and I modified from stdio to sse in L9 then I connect to the server successfully but noting just a UI… … --- ## [Creating an MCP client I got an error](https://community.deeplearning.ai/t/creating-an-mcp-client-i-got-an-error/818296)
**Author:** [@LayMui](https://community.deeplearning.ai/u/LayMui)\ **Replies:** 3\ **Last updated:** [June 13, 2025, 3:38am UTC](https://community.deeplearning.ai/t/creating-an-mcp-client-i-got-an-error/818296 "2025-06-13T03:38:34Z")
Query: Hi Error: “Could not resolve authentication method. Expected either api\_key or auth\_token to be set. Or for one of the X-Api-Key or Authorization headers to be explicitly omitted” what is missing in my setup? I … --- ## [What is the role of the Inspector Proxy address?](https://community.deeplearning.ai/t/what-is-the-role-of-the-inspector-proxy-address/821593)
**Author:** [@simorgh10](https://community.deeplearning.ai/u/simorgh10)\ **Replies:** 1\ **Last updated:** [June 2, 2025, 6:24am UTC](https://community.deeplearning.ai/t/what-is-the-role-of-the-inspector-proxy-address/821593 "2025-06-02T06:24:52Z")
Hello, I’m wondering what task does the Inspector Proxy fulfill in the lab setting ? Why do we need it ? Thank you, --- ## [Is there anyone knows other MCP Servers Deployment Methods that is easier to deploy?](https://community.deeplearning.ai/t/is-there-anyone-knows-other-mcp-servers-deployment-methods-that-is-easier-to-deploy/825178)
**Author:** [@XianJialu](https://community.deeplearning.ai/u/XianJialu)\ **Replies:** 0\ **Last updated:** [May 31, 2025, 9:33pm UTC](https://community.deeplearning.ai/t/is-there-anyone-knows-other-mcp-servers-deployment-methods-that-is-easier-to-deploy/825178 "2025-05-31T21:33:06Z")
Hi, I’m currently learning about MCP Server deployment using Render and Cloudflare, but I’m curious: are there other lightweight or innovative approaches that are easier or faster to deploy? as easy as Vercel? And also… [Next page](https://community.deeplearning.ai/c/short-course-q-a/mcp-build-rich-context-ai-apps-with-anthropic/525.md?page=1)