Interactive search results

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

Annotations only declare readOnlyHint=true; the description adds the behavior that actually matters: synchronous polling of every indexer, 15s+ typical, 120s timeout, and an explicit no-retry instruction. It further discloses refusal behavior (`season` refused on Radarr, `episode` mutually exclusive with `season`) and a data quirk (`seeders` absent, not zero, on usenet) that no other field carries.

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?

Purpose is front-loaded in the first sentence, followed by differentiation, then the slow/timeout warning and per-service narrowing rules — a sensible order. The prose is dense but nearly every clause carries a distinct, actionable fact; the bolded '**Slow**' aside and the field-reference pointer at the end are slightly crammed rather than cleanly separated.

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 10-parameter, five-service tool, this covers the things an agent would otherwise get wrong: latency and retry policy, service-specific parameter refusals, breadth controls, and the book path via shelfmark. Return shape is partly covered by the existing output schema, so the description is not obligated to explain values, and it still explains what `rejections` means and where `guid` lives.

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 already 90%, so the baseline is 3; the description earns above that by adding cross-parameter meaning the schema lacks — that Lidarr's `id` is the artist so `album` should be passed to avoid thousand-row responses, that `episode` is far faster than `season`, and that `detail: "full"` is what surfaces the `guid`/`indexerId` `grab_release` needs. It does not restate schema text, but it also doesn't touch `limit`, `offset`, or `instance`, leaving a small gap.

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 — runs the same interactive indexer search the *arr UI runs — and scopes the output precisely (every candidate, rejected ones included, with `rejections`). It explicitly contrasts itself with `trigger_search` and names `grab_release` as the consumer of its `guid`/`indexerId`, so an agent can place it in the workflow without opening 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?

Gives explicit routing rules: use this instead of `trigger_search` when you need to see what was found, keep searches narrow (episode over season on Sonarr, album over artist on Lidarr), and use `service: "shelfmark"` for books. It also includes a when-not rule — 'a long wait is not a hang: do not retry' — which is exactly the kind of exclusion agents otherwise get wrong.

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