zdx_list_application_users

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

Annotations already declare readOnlyHint=true and openWorldHint=false, and the description's own 'Read-only' is consistent (no annotation contradiction). Beyond the annotations, the description discloses meaningful behavioral traits: results are 'curated rows' (not raw API output), the default `since` window is 2 hours, and the row shape (id, name, email, ZDX score). This adds behavioral value beyond what the structured fields provide, though it could also disclose rate limits or ordering.

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?

Four tight sentences, front-loaded with the core purpose and immediately valuable details. No wasted words, and the structure flows naturally from what → characteristics → filters → next step. The 'Read-only.' sentence is slightly redundant with annotations but doesn't detract.

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 list tool with 7 parameters and no output schema, the description covers the essentials: return shape, filters, default window, and chaining. It communicates what a returned row contains and how to drill into a single user, which is the key workflow a caller needs. It could mention pagination, timezone semantics, or auth requirements, but the '%s' curated-rows note plus the defaults cover the most error-prone areas.

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 low at 14% (only `query` has its own description), so the description must carry the load. It does so effectively by explaining `score_bucket` values (poor/okay/good), the `since` unit (HOURS, default 2h), and the filter dimensions. The `query` parameter has an exceptional description about JMESPath, snake_case gotchas, and a safety tip to call without `query` first to read keys. However, `app_id`, `location_id`, `department_id`, and `geo_id` get no dedicated semantic detail beyond a name mention.

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 uses a specific verb+resource: 'List users/devices that accessed a ZDX application' with the critical qualifier 'as curated rows.' It clarifies the unit of output ('one triage row per user (id, name, email, ZDX score)'), which distinguishes it from sibling tools like `zdx_list_devices` or `zdx_list_applications`. The chaining reference to `zdx_get_application_user` further disambiguates it from the singular getter variant.

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 tells the agent exactly when to use it and how: filter by `score_bucket` (with enumerated values poor/okay/good), location/department/geo, and the `since` HOURS window (default 2h). It gives a direct workflow instruction to use a returned `id` with `zdx_get_application_user`, which functions as guidance for next steps. It stops short of explicitly listing when-not-to-use situations, but the filter and chaining guidance give clear operational context.

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