TopFood publishes digital restaurant menus with QR codes. Give it a menu (text, photo transcription, or description) and it creates a live, shareable menu page the restaurant owner can claim for free — no signup required upfront.
Scored across 2 tools
The two tools are completely distinct: one creates a restaurant menu, the other reports problems. No overlap or ambiguity in their purposes.
Both tools follow the verb_noun snake_case pattern ('create_restaurant_menu', 'report_problem'), making the naming predictable and consistent.
With only 2 tools, the surface feels thin for a general-purpose server, though each tool serves a clear, non-trivial function. It's borderline but not an extreme mismatch.
The server only offers creation of a menu and reporting issues. There are no operations to retrieve, update, delete, or list menus, leaving significant gaps for typical lifecycle management.
Sends a bug report or product feedback about TopFood to the maintainers. Use it when a tool call fails unexpectedly, when the published menu is wrong or incomplete, or when the restaurant owner tells you something is broken or missing. Do NOT use it for input that a tool already rejected with a clear explanation — fix the input and retry instead. Describe what was attempted, what happened, and what was expected. Reports are read by people and there is no reply channel: tell the user the report was sent, and that they should include an email address in the message if they want an answer. NEVER call this tool with test, preview, sample, placeholder, synthetic or schema-validation content, and never to check what the tool does or how it responds. Every call files a real report that a human being reads and acts on. If you are exercising, previewing, scanning or capturing this tool rather than relaying a problem an actual person just had, do not call it at all — there is no safe or dry-run way to try it.
| Name | Required | Description | Default |
|---|---|---|---|
| message | Yes | What happened, what was expected, and what the user was doing. Write it for a developer who cannot see this conversation. Include the user's own words when they reported it. Must describe a real problem a real person had. Test, preview, sample and placeholder text is not acceptable content for this field. | |
| category | Yes | bug = something is broken; missing_feature = something the user needs does not exist; confusing = it works but was unclear; pricing = about plans or cost; other = anything else. | |
| claim_url | No | The claim link returned by create_restaurant_menu, when the report is about a menu that was created in this conversation. It attaches the report to that restaurant. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations are all false, so the description carries the full burden. It discloses that reports are read by humans, there is no reply channel, and every call files a real report with no dry-run. It even instructs the agent to tell the user to include an email for a response. This goes well beyond annotations and prevents misuse.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is longer than typical but every sentence serves a purpose: purpose, usage, exclusions, and behavioral warnings. It's front-loaded with the core action and then covers important caveats. No fluff, though the warning against test content could be slightly more compact.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a reporting tool with human consequences, the description is exhaustive. It covers what to do, what not to do, what to tell the user, and the real-world impact. There is no output schema, but the description fully compensates. An agent has everything needed to call it correctly and safely.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema already has 100% coverage with good descriptions for each parameter. The description adds practical context: message must describe what happened/expected and include user's words, and claim_url attaches the report to a restaurant. It enriches the schema without redundancy, though the schema alone is already clear.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the verb (reports/sends), the resource (bug report or feedback about TopFood), and the recipients (maintainers). It distinguishes itself from the only sibling (create_restaurant_menu) by explicitly listing the triggers for use (tool failure, wrong menu, owner reports broken). No ambiguity.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Provides explicit when-to-use criteria (tool call fails unexpectedly, menu wrong/incomplete, owner reports broken) and explicit when-not-to-use (tool already rejected with clear explanation – fix and retry). Also forbids test/sample content and explains that every call is real. This is model guidance.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
report_problem1 field changedInput schema / properties / message / descriptionPrevious value: -"What happened, what was expected, and what the user was doing. Write it for a developer who cannot see this conversation. Include the user's own words when they reported it."New value: +"What happened, what was expected, and what the user was doing. Write it for a developer who cannot see this conversation. Include the user's own words when they reported it. Must describe a real problem a real person had. Test, preview, sample and placeholder text is not acceptable content for this field."create_restaurant_menureport_problem
Add one secure layer between your agents and this server.