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

The description goes beyond the annotations by disclosing the matching algorithm (greedy matching at IoU threshold, highest-confidence first, each ground-truth box matched at most once), the interpolation method (all-points interpolation used by Pascal VOC 2010 onward), and the error handling behavior (never raises a protocol error, returns an error object). It also confirms the read-only and idempotent nature, which aligns with the annotations. This adds significant behavioral context beyond what the annotations provide.

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 description is well-structured and front-loaded. It starts with a clear one-sentence purpose, then explains the algorithm, provides a concrete example, gives usage guidance, and ends with error handling. Every sentence adds value, and the structure is logical. It is concise yet comprehensive, with no wasted words.

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?

Given the tool's complexity (6 parameters, matching algorithm, error handling), the description is remarkably complete. It covers the purpose, algorithm, input format, output format, usage guidance, error behavior, and safety (read-only, idempotent). The output schema exists, so the description doesn't need to explain return values in detail, but it provides a concrete example. This is a fully self-contained description that leaves no major gaps.

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?

The schema description coverage is 100%, so the schema already documents all parameters. The description adds value by explaining the typical input format and the default behavior (e.g., score defaults to 1.0 when omitted, iou_threshold default 0.5). It also clarifies the box_format parameter by mentioning the default 'voc' and that image_width/height are required for yolo boxes. However, it doesn't add much beyond what the schema already states, so a 4 is appropriate.

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 clearly states the tool's purpose: 'Score detections against ground truth and show the working.' It specifies the exact operation (scoring detections against ground truth) and the resource (detections and ground truth). It distinguishes itself from siblings by explicitly noting it is not for cleaning up overlapping output (that's nms) and by mentioning the PREMIUM license, which sets it apart from other tools in the set.

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 provides explicit usage guidance: 'Use to compare two models on the same held-out set. Not for cleaning up a single model's overlapping output first - run nms before this.' This clearly states when to use the tool and when not to, and even names the alternative tool (nms). This is a model example of usage guidelines.

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