deliveroo_search

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations provided, the description carries the full burden of behavioral disclosure. It discloses that the tool returns the entire feed (not curated), that results can be up to 2,700+, that each result carries a uname for downstream endpoints, and that unrecognized filter values are silently ignored. It does not explicitly state whether the tool is read-only, but the description's emphasis on search and the absence of mutation language make the read-only nature clear enough. A small gap: it doesn't mention rate limits or pagination, but the description is unusually rich for a search tool.

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 dense but well-organized: it front-loads the core purpose, then explains the feed scope, result fields, filter behavior, and edge case. Every sentence adds information, and the reference to deliveroo_search_filters is placed where it's most relevant. It is longer than the minimum, but the length is justified by the tool's 15 parameters and the need to explain the uname contract and filter semantics.

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 search tool with 15 parameters and no output schema, the description covers the essential context: what the tool returns, how to get valid filter values, how to use the uname for downstream calls, and the empty-list edge case. It doesn't describe the exact response structure (e.g., whether results are nested under a 'data' key), but the description's mention of result fields partially compensates. The absence of an output schema makes this a notable gap, but the description is still more complete than most.

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 description coverage is 100%, so the schema already documents all 15 parameters. The description adds value by explaining the collection parameter's vertical selection (restaurants/grocery/shopping), the top_rated equivalence to min_rating=4.5, and the fact that unrecognized filter values are silently ignored. It also explains the uname contract for downstream endpoints. This goes beyond the schema's terse descriptions, though it doesn't add syntax details for every parameter.

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 opens with a specific verb ('Search') and resource ('Deliveroo restaurants, groceries, or shopping stores near a location'), and immediately distinguishes the tool's scope from curated subsets by noting it returns the location's entire feed. It also names the sibling deliveroo_search_filters as the companion for valid filter values, which helps an agent tell this tool apart from related search tools.

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?

The description explicitly states when to use this tool (searching Deliveroo's catalog by lat/long) and when to use the sibling deliveroo_search_filters (to get valid cuisine/dietary/dish values). It also clarifies that a location with no coverage returns an empty list rather than an error, which is a useful behavioral cue for handling results.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.