# MCP Course: Error when running tiny-agents example **URL:** \ **Category:** Course\ **Created:** [December 29, 2025, 8:09pm UTC](https://discuss.huggingface.co/t/mcp-course-error-when-running-tiny-agents-example/172083 "2025-12-29T20:09:02Z")\ **Posts on this page:** 4\ **Page:** 1
**Author:** ![krosenke](https://avatars.discourse-cdn.com/v4/letter/k/b5a626/32.png) [@krosenke](https://discuss.huggingface.co/u/krosenke)\ **Post date:** [December 29, 2025, 8:09pm UTC](https://discuss.huggingface.co/t/mcp-course-error-when-running-tiny-agents-example/172083/1 "2025-12-29T20:09:02Z")
Error when running tiny-agents example I’m following the MCP course and I’m stuck at the end of the MCP Clients section. When running the command “tiny-agents run agent.json” I get the error below. I followed all instructions. What am I missing? An unexpected error occurred: ‘config’ Traceback (most recent call last): File “C:\Python312\Lib\site-packages\huggingface\_hub\inference\_mcp\cli.py”, line 134, in run\_agent await agent.load\_tools() File “C:\Python312\Lib\site-packages\huggingface\_hub\inference\_mcp\agent.py”, line 57, in load\_tools await self.add\_mcp\_server(cfg[“type”], \*\*cfg[“config”]) ~~~^^^^^^^^^^ KeyError: ‘config’ An unexpected error occurred: ‘config’ ╭───────────────────────────────────────── Traceback (most recent call last) ──────────────────────────────────────────╮ │ C:\Python312\Lib\site-packages\huggingface\_hub\inference\_mcp\cli.py:225 in run │ │ │ │ 222 │ │ raise typer.Exit(code=130) │ │ 223 │ except Exception as e: │ │ 224 │ │ print(f"\n[bold red]An unexpected error occurred: {e}[/bold red]“, flush=True) │ │ ❱ 225 │ │ raise e │ │ 226 │ │ 227 │ │ 228 if **name** == “ **main** ”: │ │ │ │ ╭────── locals ───────╮ │ │ │ path = ‘agent.json’ │ │ │ ╰─────────────────────╯ │ │ │ │ C:\Python312\Lib\site-packages\huggingface\_hub\inference\_mcp\cli.py:219 in run │ │ │ │ 216 │ ), │ │ 217 ): │ │ 218 │ try: │ │ ❱ 219 │ │ asyncio.run(run\_agent(path)) │ │ 220 │ except KeyboardInterrupt: │ │ 221 │ │ print(”\n[red]Application terminated by KeyboardInterrupt.[/red]“, flush=True) │ │ 222 │ │ raise typer.Exit(code=130) │ │ │ │ ╭────── locals ───────╮ │ │ │ path = ‘agent.json’ │ │ │ ╰─────────────────────╯ │ │ │ │ C:\Python312\Lib\asyncio\runners.py:194 in run │ │ │ │ 191 │ │ │ “asyncio.run() cannot be called from a running event loop”) │ │ 192 │ │ │ 193 │ with Runner(debug=debug, loop\_factory=loop\_factory) as runner: │ │ ❱ 194 │ │ return runner.run(main) │ │ 195 │ │ 196 │ │ 197 def \_cancel\_all\_tasks(loop): │ │ │ │ ╭─────────────────────────────── locals ───────────────────────────────╮ │ │ │ debug = None │ │ │ │ loop\_factory = None │ │ │ │ main = \ │ │ │ │ runner = \ │ │ │ ╰──────────────────────────────────────────────────────────────────────╯ │ │ │ │ C:\Python312\Lib\asyncio\runners.py:118 in run │ │ │ │ 115 │ │ │ │ 116 │ │ self.\_interrupt\_count = 0 │ │ 117 │ │ try: │ │ ❱ 118 │ │ │ return self.\_loop.run\_until\_complete(task) │ │ 119 │ │ except exceptions.CancelledError: │ │ 120 │ │ │ if self.\_interrupt\_count \> 0: │ │ 121 │ │ │ │ uncancel = getattr(task, “uncancel”, None) │ │ │ │ ╭───────────────────────────────────────────────────── locals ─────────────────────────────────────────────────────╮ │ │ │ context = \<\_contextvars.Context object at 0x000001F739D12840\> │ │ │ │ coro = \ │ │ │ │ self = \ │ │ │ │ sigint\_handler = functools.partial(\\>, main\_task=\ │ │ │ │ exception=KeyError(‘config’)\>) │ │ │ │ task = \ │ │ │ │ exception=KeyError(‘config’)\> │ │ │ ╰──────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯ │ │ │ │ C:\Python312\Lib\asyncio\base\_events.py:687 in run\_until\_complete │ │ │ │ 684 │ │ if not future.done(): │ │ 685 │ │ │ raise RuntimeError(‘Event loop stopped before Future completed.’) │ │ 686 │ │ │ │ ❱ 687 │ │ return future.result() │ │ 688 │ │ │ 689 │ def stop(self): │ │ 690 │ │ “”“Stop running the event loop. │ │ │ │ ╭───────────────────────────────────────────────────── locals ─────────────────────────────────────────────────────╮ │ │ │ future = \ │ │ │ │ exception=KeyError(‘config’)\> │ │ │ │ new\_task = False │ │ │ │ self = │ │ │ ╰──────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯ │ │ │ │ C:\Python312\Lib\site-packages\huggingface\_hub\inference\_mcp\cli.py:194 in run\_agent │ │ │ │ 191 │ except Exception as e: │ │ 192 │ │ tb\_str = traceback.format\_exc() │ │ 193 │ │ print(f”\n[bold red]An unexpected error occurred: {e}\n{tb\_str}[/bold red]”, flu │ │ ❱ 194 │ │ raise e │ │ 195 │ │ │ 196 │ finally: │ │ 197 │ │ if sigint\_registered\_in\_loop: │ │ │ │ ╭───────────────────────────────────────────────────── locals ─────────────────────────────────────────────────────╮ │ │ │ abort\_event = \ │ │ │ │ agent = \ │ │ │ │ agent\_path = ‘agent.json’ │ │ │ │ config = { │ │ │ │ │ ‘model’: ‘Qwen/Qwen2.5-72B-Instruct’, │ │ │ │ │ ‘provider’: ‘nebius’, │ │ │ │ │ ‘servers’: [ │ │ │ │ │ │ {‘type’: ‘stdio’, ‘command’: ‘npx’, ‘args’: [‘@playwright/mcp@latest’]} │ │ │ │ │ ] │ │ │ │ } │ │ │ │ exit\_event = \ │ │ │ │ first\_sigint = True │ │ │ │ inputs = │ │ │ │ loop = │ │ │ │ original\_sigint\_handler = functools.partial(\\>, main\_task=\ │ │ │ │ exception=KeyError(‘config’)\>) │ │ │ │ prompt = None │ │ │ │ servers = [{‘type’: ‘stdio’, ‘command’: ‘npx’, ‘args’: [‘@playwright/mcp@latest’]}] │ │ │ │ sigint\_registered\_in\_loop = False │ │ │ │ tb\_str = ‘Traceback (most recent call last):\n File │ │ │ │ “C:\Python312\Lib\site-packag’+346 │ │ │ ╰──────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯ │ │ │ │ C:\Python312\Lib\site-packages\huggingface\_hub\inference\_mcp\cli.py:134 in run\_agent │ │ │ │ 131 │ │ │ servers=servers, │ │ 132 │ │ │ prompt=prompt, │ │ 133 │ │ ) as agent: │ │ ❱ 134 │ │ │ await agent.load\_tools() │ │ 135 │ │ │ print(f”[bold blue]Agent loaded with {len(agent.available\_tools)} tools:[/bo │ │ 136 │ │ │ for t in agent.available\_tools: │ │ 137 │ │ │ │ print(f"[blue] • {t.function.name}[/blue]") │ │ │ │ ╭───────────────────────────────────────────────────── locals ─────────────────────────────────────────────────────╮ │ │ │ abort\_event = \ │ │ │ │ agent = \ │ │ │ │ agent\_path = ‘agent.json’ │ │ │ │ config = { │ │ │ │ │ ‘model’: ‘Qwen/Qwen2.5-72B-Instruct’, │ │ │ │ │ ‘provider’: ‘nebius’, │ │ │ │ │ ‘servers’: [ │ │ │ │ │ │ {‘type’: ‘stdio’, ‘command’: ‘npx’, ‘args’: [‘@playwright/mcp@latest’]} │ │ │ │ │ ] │ │ │ │ } │ │ │ │ exit\_event = \ │ │ │ │ first\_sigint = True │ │ │ │ inputs = │ │ │ │ loop = │ │ │ │ original\_sigint\_handler = functools.partial(\\>, main\_task=\ │ │ │ │ exception=KeyError(‘config’)\>) │ │ │ │ prompt = None │ │ │ │ servers = [{‘type’: ‘stdio’, ‘command’: ‘npx’, ‘args’: [‘@playwright/mcp@latest’]}] │ │ │ │ sigint\_registered\_in\_loop = False │ │ │ │ tb\_str = ‘Traceback (most recent call last):\n File │ │ │ │ "C:\Python312\Lib\site-packag’+346 │ │ │ ╰──────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯ │ │ │ │ C:\Python312\Lib\site-packages\huggingface\_hub\inference\_mcp\agent.py:57 in load\_tools │ │ │ │ 54 │ │ │ 55 │ async def load\_tools(self) → None: │ │ 56 │ │ for cfg in self.\_servers\_cfg: │ │ ❱ 57 │ │ │ await self.add\_mcp\_server(cfg[“type”], \*\*cfg[“config”]) │ │ 58 │ │ │ 59 │ async def run( │ │ 60 │ │ self, │ │ │ │ ╭───────────────────────────────────── locals ─────────────────────────────────────╮ │ │ │ cfg = {‘type’: ‘stdio’, ‘command’: ‘npx’, ‘args’: [‘@playwright/mcp@latest’]} │ │ │ │ self = \ │ │ │ ╰──────────────────────────────────────────────────────────────────────────────────╯ │ ╰──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯ KeyError: ‘config’ ---
**Author:** ![John6666](https://sea2.discourse-cdn.com/hellohellohello/user_avatar/discuss.huggingface.co/john6666/32/27664_2.png) [@John6666](https://discuss.huggingface.co/u/John6666)\ **Post date:** [December 29, 2025, 11:42pm UTC](https://discuss.huggingface.co/t/mcp-course-error-when-running-tiny-agents-example/172083/2 "2025-12-29T23:42:04Z")
The probable cause seems to be this. * * * ### Root cause You are running a **Python** Tiny Agents runner (traceback is in `huggingface_hub ... inference_mcp ... cli.py`). That runner iterates `servers` and does `cfg["config"]`. Your `agent.json` has a server entry with **no** `"config"` key, so it crashes with `KeyError: 'config'`. This happens because the HF course and HF JS docs show a **different** (flat) server shape that is valid for the **JS** CLI (`npx @huggingface/tiny-agents ...`), while some **Python** runners expect a nested `"config"` object. You can see the course page showing a flat server entry (`command` and `args` at the top level) and telling you to run with `npx @huggingface/tiny-agents run agent.json`. ([Hugging Face](https://huggingface.co/learn/mcp-course/en/unit2/tiny-agents "Building Tiny Agents with MCP and the Hugging Face Hub - Hugging Face MCP Course")) You can also see the JS tiny-agents docs showing the same flat shape and the same `npx @huggingface/tiny-agents run ...` usage. ([Hugging Face](https://huggingface.co/docs/huggingface.js/en/tiny-agents/README "@huggingface/tiny-agents")) Separately, you can see configs that use the nested `"config"` shape in the wild (example: AMD’s guide uses `"config": {"command": "...npx.cmd", "args": ["-y", ...]}`) while still running `tiny-agents run`. ([AMD](https://www.amd.com/ja/developer/resources/technical-articles/2025/local-tiny-agents--mcp-agents-on-ryzen-ai-with-lemonade-server.html "Local Tiny Agents: MCP Agents on Ryzen AI with Lemonade Server")) * * * ## Why there are two shapes (background) ### Shape A. “Flat server object” (common in JS tiny-agents examples) Used in: - HF MCP course “Tiny Agents MCP Client in the Command Line” example JSON. ([Hugging Face](https://huggingface.co/learn/mcp-course/en/unit2/tiny-agents "Building Tiny Agents with MCP and the Hugging Face Hub - Hugging Face MCP Course")) - HF JS tiny-agents docs example JSON. ([Hugging Face](https://huggingface.co/docs/huggingface.js/en/tiny-agents/README "@huggingface/tiny-agents")) Looks like: ```json { "type": "stdio", "command": "npx", "args": ["@playwright/mcp@latest"] } ``` ### Shape B. “Nested config” (some Python runners and other guides) Used in: - AMD guide example JSON (explicitly shows `config` nesting and Windows `npx.cmd`). ([AMD](https://www.amd.com/ja/developer/resources/technical-articles/2025/local-tiny-agents--mcp-agents-on-ryzen-ai-with-lemonade-server.html "Local Tiny Agents: MCP Agents on Ryzen AI with Lemonade Server")) - HF Hub issue reports also discuss nested `config` being passed around and causing KeyErrors depending on the library version. ([GitHub](https://github.com/huggingface/huggingface_hub/issues/3203 "KeyError: 'command' in add\_mcp\_server when running tiny-agent due to incorrect params structure · Issue #3203 · huggingface/huggingface\_hub · GitHub")) Looks like: ```json { "type": "stdio", "config": { "command": "npx", "args": ["@playwright/mcp@latest"] } } ``` ### Why it’s confusing The course page is explicitly teaching the **JS** CLI path (`npx @huggingface/tiny-agents ...`) and shows the flat JSON. ([Hugging Face](https://huggingface.co/learn/mcp-course/en/unit2/tiny-agents "Building Tiny Agents with MCP and the Hugging Face Hub - Hugging Face MCP Course")) But your command `tiny-agents run agent.json` is executing a **Python** -installed command (based on the traceback path). That Python runner expects the nested key and fails. * * * ## Fix 1 (most direct): add `"config"` in your `agent.json` Replace your server entry: ```json { "type": "stdio", "command": "npx", "args": ["@playwright/mcp@latest"] } ``` With: ```json { "type": "stdio", "config": { "command": "npx", "args": ["@playwright/mcp@latest"] } } ``` Full example matching your model/provider: ```json { "model": "Qwen/Qwen2.5-72B-Instruct", "provider": "nebius", "servers": [ { "type": "stdio", "config": { "command": "npx", "args": ["@playwright/mcp@latest"] } } ] } ``` ### Windows-hardening tweaks (strongly recommended) On Windows, it is often more reliable to point to `npx.cmd` explicitly and disable prompts: AMD’s working Windows example uses a full path to `npx.cmd` and passes `-y` in args. ([AMD](https://www.amd.com/ja/developer/resources/technical-articles/2025/local-tiny-agents--mcp-agents-on-ryzen-ai-with-lemonade-server.html "Local Tiny Agents: MCP Agents on Ryzen AI with Lemonade Server")) So you can do: ```json { "type": "stdio", "config": { "command": "C:\\Program Files\\nodejs\\npx.cmd", "args": ["-y", "@playwright/mcp@latest"] } } ``` * * * ## Fix 2 (keep the course JSON): run the JS CLI exactly as the course says If you want to keep the **flat** server JSON as shown in the course, run it the way the course instructs: - `npx @huggingface/tiny-agents run agent.json` ([Hugging Face](https://huggingface.co/learn/mcp-course/en/unit2/tiny-agents "Building Tiny Agents with MCP and the Hugging Face Hub - Hugging Face MCP Course")) This matches the flat schema shown in both the course and the JS docs. ([Hugging Face](https://huggingface.co/learn/mcp-course/en/unit2/tiny-agents "Building Tiny Agents with MCP and the Hugging Face Hub - Hugging Face MCP Course")) * * * ## Version drift is real (why you see KeyErrors at all) Hugging Face has a known class of issues where the runner expects one structure but receives the other. Example: a Hub issue report shows `KeyError: 'command'` when a nested `config` dict is provided but the underlying code expects `params["command"]` at the top level. ([GitHub](https://github.com/huggingface/huggingface_hub/issues/3203 "KeyError: 'command' in add\_mcp\_server when running tiny-agent due to incorrect params structure · Issue #3203 · huggingface/huggingface\_hub · GitHub")) Your case is the mirror image: the runner expects `cfg["config"]`, but your JSON is flat. This is why two people can “follow the docs” and still hit different KeyErrors. * * * ## Quick sanity checks (so you stop guessing) ### 1) Confirm which `tiny-agents` you are running Your traceback indicates Python, but you can confirm by checking where the executable resolves: - Windows: `where tiny-agents` If it points into a Python Scripts directory, you are on the Python runner. ### 2) Confirm you installed the course’s prerequisites The course explicitly says you need Node tooling (`npm` and `npx`) even if you are using Python, and it shows installing `huggingface_hub[mcp]` with a minimum version. ([Hugging Face](https://huggingface.co/learn/mcp-course/en/unit1/mcp-clients "MCP Clients - Hugging Face MCP Course")) * * * ## High-quality references (keep these open while debugging) - HF MCP Course Unit 2 Tiny Agents page (shows flat JSON + `npx @huggingface/tiny-agents run ...`). [https://huggingface.co/learn/mcp-course/en/unit2/tiny-agents](https://huggingface.co/learn/mcp-course/en/unit2/tiny-agents) ([Hugging Face](https://huggingface.co/learn/mcp-course/en/unit2/tiny-agents "Building Tiny Agents with MCP and the Hugging Face Hub - Hugging Face MCP Course")) - HF JS tiny-agents docs (flat JSON examples and CLI usage). [https://huggingface.co/docs/huggingface.js/en/tiny-agents/README](https://huggingface.co/docs/huggingface.js/en/tiny-agents/README)_toggle “Define your own agent”_ ([Hugging Face](https://huggingface.co/docs/huggingface.js/en/tiny-agents/README "@huggingface/tiny-agents")) - HF Hub MCP Python reference (defines stdio params like `command`, `args`, `env`, `cwd`). [https://huggingface.co/docs/huggingface\_hub/en/package\_reference/mcp](https://huggingface.co/docs/huggingface_hub/en/package_reference/mcp) ([Hugging Face](https://huggingface.co/docs/huggingface_hub/en/package_reference/mcp "MCP Client")) - HF Hub issue illustrating schema mismatch leading to KeyErrors (`KeyError: 'command'`). [https://github.com/huggingface/huggingface\_hub/issues/3203](https://github.com/huggingface/huggingface_hub/issues/3203) ([GitHub](https://github.com/huggingface/huggingface_hub/issues/3203 "KeyError: 'command' in add\_mcp\_server when running tiny-agent due to incorrect params structure · Issue #3203 · huggingface/huggingface\_hub · GitHub")) - AMD guide showing a working Windows nested `config` + `npx.cmd` + `-y` pattern. [https://www.amd.com/ja/developer/resources/technical-articles/2025/local-tiny-agents–mcp-agents-on-ryzen-ai-with-lemonade-server.html](https://www.amd.com/ja/developer/resources/technical-articles/2025/local-tiny-agents--mcp-agents-on-ryzen-ai-with-lemonade-server.html) ([AMD](https://www.amd.com/ja/developer/resources/technical-articles/2025/local-tiny-agents--mcp-agents-on-ryzen-ai-with-lemonade-server.html "Local Tiny Agents: MCP Agents on Ryzen AI with Lemonade Server")) * * * ### Summary - Your error is exactly: runner expects `servers[*].config`, your JSON does not have it. - Fix path A: nest `command` and `args` under `"config"`. - Fix path B: keep flat JSON and run with `npx @huggingface/tiny-agents ...` as the course shows. - On Windows, prefer `npx.cmd` and `-y` for reliability. ---
**Author:** ![krosenke](https://avatars.discourse-cdn.com/v4/letter/k/b5a626/32.png) [@krosenke](https://discuss.huggingface.co/u/krosenke)\ **Post date:** [December 31, 2025, 4:46pm UTC](https://discuss.huggingface.co/t/mcp-course-error-when-running-tiny-agents-example/172083/3 "2025-12-31T16:46:09Z")
Thank you for the detailed explanation! It helped me to resolve the issue. I used option 1 and added “config” to the json file. **Notes:** - I got this error and fixed it by changing the provider to novita: `Error during agent run: Model Qwen/Qwen2.5-72B-Instruct is not supported by provider nebius.` - There is no video on page [https://huggingface.co/learn/mcp-course/en/unit1/mcp-clients#setup](https://huggingface.co/learn/mcp-course/en/unit1/mcp-clients#setup); only a picture is displayed. ---
**Author:** ![Robeee](https://avatars.discourse-cdn.com/v4/letter/r/3be4f8/32.png) [@Robeee](https://discuss.huggingface.co/u/Robeee)\ **Post date:** [January 13, 2026, 5:20am UTC](https://discuss.huggingface.co/t/mcp-course-error-when-running-tiny-agents-example/172083/4 "2026-01-13T05:20:20Z")
Hi there if you’re still stuck with this error I just overcame it by searching for what models does nebius support and got this huggingface link: [Models available on Nebius AI – Hugging Face](https://huggingface.co/models?inference_provider=nebius) The model i switched to was `Qwen/Qwen3-30B-A3B-Instruct-2507` and it fixed the error for me. All the best