preview

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

With no annotations, the description carries the full burden and does substantial work: it discloses async behavior on Windows, platform-specific export requirements, id invalidation on photo/layer/revision change, and retention limits (one preview job, eight PNG cache files). It stops short of saying whether kind=render writes files to disk or what side effects/restore_destination entail.

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?

Roughly a hundred words with no wasted filler, but the structure is poor: it opens with a prerequisite rather than the purpose, then interleaves platform notes, per-kind behavior, and coordinate math in run-on sentences with hard-wrapped fragments. Front-loading what the tool does would fix most of it.

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 9-parameter, nested-target, no-output-schema tool, the description covers the highest-risk semantics: async behavior, id invalidation, cache retention, and coordinate mapping. Gaps remain around the opaque `target`/`observation_token` contract and unexplained actions, which an agent would have to guess at.

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 0%, so the description must compensate, and it partly does: `kind` modes, `max_size` range 128..1600, `points` as pixel-edge {x,y} coordinates with the RAW normalization trick, `preview_id` reuse/invalidation, and `direction` implied via the divide-by-raw_dimensions note. `target`, `request_id`, `observation_token`, and three of five `action` values get no explanation at all.

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 makes clear this tool produces previews/renders of Capture One content across several modes (native adjustments, layer masks, JPEG exports, display capture), and it names the `context` sibling as a prerequisite. It never states the purpose in one declarative sentence, though — it opens with a directive ("Check context capabilities first") and lets the resource be inferred from the per-kind notes.

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?

There is real routing guidance: check `context` first, macOS requires explicitly enabled kind=render, map_points must reuse the same kind/max_size and preview_id. But there is no guidance on when to choose `preview` over siblings like `read`, `images`, or `adjust`, and the `action` values prepare/render/restore_destination are never explained, so mode selection is only half-covered.

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