Detect UI elements in a screenshot

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

With no annotations provided, the description carries the full disclosure burden and discharges it thoroughly: precision limits ('expect a pixel or two, and corner radii within about a pixel'), the downscaled-analysis recovery caveat, the renderRisk >2% crop threshold, and failure modes (figmaUrl errors without FIGMA_TOKEN; font runs with missing glyphs are baked as pictures 'always with a warning'). It even discloses default font substitution (bundled Inter), which is exactly the beyond-annotations context this dimension rewards.

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 first sentence front-loads the purpose and every sentence carries technical content, so there is no fluff — but the entire description is one dense, unbroken paragraph that mixes precision notes, renderMode semantics, workflow advice, and SVG-mode detail. Bullet points or paragraph separation between imagePath mode and svgPath/figmaUrl mode would make the high-value guidance much more scannable for an agent.

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 output schema, the description must explain return values, and it does: nested element tree with rects/corner radii/fills, numbered overlay PNG, renderMode/renderRisk/semanticHint/roleHint semantics, and coordinate behavior. Minor gaps remain (exact tree JSON shape, coordinate-origin conventions, error-message formats), but for a 7-parameter dual-mode tool with zero annotations this is highly complete.

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 the baseline is 3, but the description adds substantial behavioral meaning beyond the schema: fonts gain substitution/baking semantics, svgPath gains 'vector-shape'/'vector-text' output behavior, imagePath gains precision/estimation caveats, and figmaUrl gains the roleHint mapping for Figma layer names. This exceeds what the JSON schema provides for nearly 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 states a specific verb and resource: 'Find the rectangles, text runs and images in a UI screenshot or design comp.' It differentiates from siblings by declaring 'This is geometry only — it does not know a button from a card' and by routing the output to riv_ui_prototype, which separates it from semantic/import/editing siblings like riv_import_svg, riv_critique, and riv_edit.

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 gives explicit mode-selection guidance: 'If you have the vector source, pass svgPath instead of imagePath' and explains when figmaUrl is viable (requires FIGMA_TOKEN). It also states the follow-on workflow ('pass the result to riv_ui_prototype') and the scope boundary (geometry only). It does not explicitly exclude or compare siblings by name beyond riv_ui_prototype, which costs the fifth point.

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