Dukkani — Saudi grocery directory

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

Annotations already declare readOnlyHint, idempotentHint and openWorldHint=false, so the agent knows this is a safe, repeatable, self-contained read. The description contributes only the fact that responses include 'links to try it', with no detail on depth, freshness, language of the returned content, or whether the topic changes the shape of the answer.

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?

Two sentences, no padding, and the usage trigger is placed second where it is easy to find. The first sentence is a dense run-on parenthetical list that reads more like marketing copy than structured guidance, which hurts scannability.

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 single-optional-parameter informational tool with no output schema, the description covers what the tool is and roughly what comes back (facts plus links). It stops short of saying whether the answer is static or fetched, in which language, or what each topic returns, which leaves real gaps for a content-serving tool.

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 0% and the description never uses the actual enum tokens. The parenthetical list (cashier, debt book, barcode, worker, orders) loosely maps to some enum values but omits 'overview' and 'einvoice' entirely, so an agent must still infer which topic string to send for an e-invoicing question.

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 first sentence states a concrete resource (facts about the Dukkani product) and enumerates the covered domains: cashier, debt book, barcode, worker language, neighbourhood orders. It is clearly an informational lookup rather than an action, which separates it from siblings like prepare_grocery_order, but it never explicitly names or contrasts those siblings.

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 second sentence gives a real trigger condition: 'Use when a shop owner asks for a cashier/POS app, a debt book app or how to run a grocery from a phone in Saudi Arabia.' That is clear when-to-use guidance, but there is no exclusion (e.g. when to prefer find_groceries or list_cities instead) and no mention of how the topic parameter should be chosen.

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

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

Annotations already cover read-only, idempotent and closed-world semantics, so the description's added context is what counts: it discloses result ordering (on_dukkani=true first), that distance is only present 'when known', and that order_url is a Dukkani page where the customer completes the order over WhatsApp. Silent on empty-result behavior and on the limit cap of 15.

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?

Two dense sentences, front-loaded with the action and scope before the return-shape and ordering details. Every clause carries information; the only mild bloat is the parenthetical Arabic gloss on بقالات/تموينات, which is still purposeful for a Saudi-market tool.

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?

There is no output schema, and the description compensates well by naming the returned fields (name, city, distance, order_url, map_url) and the WhatsApp ordering flow. What remains uncovered is failure/empty-result behavior and the meaning of on_dukkani=false rows, which for a lookup tool is a modest gap rather than a blocker.

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 coverage is 80%, so the schema already documents city, district, latitude and longitude. The description adds the one thing the schema does not: that coordinates are an alternative search mode to city/district rather than an additional filter. It says nothing about limit (default 8, max 15), leaving that entirely to the schema — baseline 3 given high coverage.

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?

States a specific verb (Find) and resource (neighbourhood groceries) with explicit source (Dukkani's public directory) and geographic scope (Saudi Arabia). The enumeration of the three search modes (district+city, city, coordinates) makes it unambiguous against siblings like prepare_grocery_order or list_cities without needing to name them.

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?

Gives clear usage context: 'Use the Arabic district/city names when you have them', and the mode list implies when each parameter combination applies. It stops short of explicit routing ('if the user wants to order, use prepare_grocery_order', 'for city discovery use list_cities'), so no exclusions or alternatives are stated.

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

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

Annotations already declare readOnly, idempotent, and closed-world, so the safety profile is covered. The description adds real context beyond them: results are limited to cities with actual listings and include a per-city count, which prevents the agent from treating this as a full gazetteer. It stops short of describing ordering or pagination, but the result set is inherently small.

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?

Two sentences, front-loaded with what the tool returns and followed by the triggering condition. No redundant restatement of the title or annotations.

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?

With no parameters and no output schema, the description carries the burden of describing the return value, and 'cities ... with how many' conveys a name-plus-count list adequately. It doesn't state the count's unit (groceries) explicitly or the ordering, a minor residual gap for a discovery tool.

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?

Zero parameters, so the baseline is 4. The description correctly signals a no-argument call by framing it as a directory lookup with no filters.

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?

States a specific verb and resource (list the Saudi cities) plus a precise scope: only cities that have groceries in Dukkani's directory, and it returns counts. This is distinct from find_groceries (which searches within a city) and prepare_grocery_order without needing to open either schema.

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?

Gives two concrete invocation triggers: a city that isn't found and a user asking where the service works. That implicitly routes the agent away from find_groceries, but no sibling is named explicitly and there is no stated 'do not use for X' exclusion.

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

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

Annotations already declare readOnlyHint/idempotent/openWorld=false; the description adds the important reassurance that 'this tool sends nothing by itself', resolving what an 'order' tool actually does. It also discloses the return payload (cleaned list + order_url), which the annotations cannot convey.

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?

Three tight sentences: action, mechanics/return, then the key non-side-effect caveat. No filler and the most decision-relevant information (id source, no sending) is front-loaded.

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 3-parameter tool with no output schema, the description covers inputs, the return shape, and the non-destructive behavior. It does not mention failure modes such as an invalid or unknown grocery_id, which would be the last missing piece.

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 coverage is 100%, so baseline is 3, but the description adds meaning: grocery_id is a foreign key from find_groceries and items are normalized ('returns the cleaned list'), telling the agent input phrasing is flexible. It does not describe note handling or item-limit behavior.

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?

States a specific verb+resource (turn a shopping list into a ready order) and names the sibling find_groceries as the required id source. An agent can distinguish this from find_groceries or dukkani_for_shop_owners without opening any schema.

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?

Explicitly says the grocery_id must come from find_groceries, giving the correct upstream call order, and clarifies the customer completes the order elsewhere. It lacks an explicit 'do not use for X' exclusion, but the when-to-use context is unambiguous.

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