kokon

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

Annotations already declare the safety profile (destructiveHint=false, idempotentHint=false), so the bar is lower. The description adds genuinely useful behavior beyond those: the input file is never overwritten and width/height/aspect ratio are preserved by default. It stops short of describing the return payload or latency, but the added mutation-safety context is solid.

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?

Three short sentences, front-loaded with purpose, then the default behavior, then the non-destructive guarantee. Every sentence carries information and nothing is padded.

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 7-parameter non-read-only tool with no output schema, the description covers purpose, preservation default, and the no-overwrite guarantee. It does not address seed/quality trade-offs or what the result contains, but those are handled by the input schema and include_preview documentation. Sufficient, with minor room to grow.

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 100%, so the baseline is 3. The description's prompt example and the dimension-preservation default merely mirror what the schema already documents, adding no syntax or semantics beyond it.

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 gives a specific verb and resource ('Edit an existing image') plus the interaction mode ('natural-language instruction') with a concrete example. This clearly separates it from siblings like generate_image (creation), inpaint_image/outpaint_image (region-specific operations), and edit_images (plural/batch).

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?

Usage is implied by 'Edit an existing image', and the default dimension-preservation behavior is stated, but the description never says when to prefer this over edit_images, inpaint_image, or outpaint_image, nor does it state prerequisites. No exclusions or routing guidance is offered.

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

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

Annotations only declare the general write/non-idempotent profile, so the description adds real value: the ordering contract ('refer to them as image 1, image 2, ... in the prompt') and the default output sizing rule ('follows image 1's aspect ratio unless width/height or aspect_ratio are given'). It does not cover file-safety behavior on output (non-overwrite, project-scoped paths), though the schema does.

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?

Three sentences, front-loaded with the core action, followed by the two rules an agent would otherwise get wrong (ordering and default aspect ratio). The only mild redundancy is '(up to 10)', which the schema's maxItems already states.

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 generation tool with no output schema, the description supplies the key invocation semantics and the include_preview parameter hints at the return shape, but it never states what a successful call yields (image path/list) or that generation is nondeterministic unless seed is set. The remaining gap is small given the strong schema coverage.

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 78%, so the schema already carries most parameter meaning (seed, quality, output_path, aspect_ratio are all documented inline). The description goes beyond it by explaining parameter interaction: reference order maps to 'image 1, image 2' labels in the prompt, and width/height/aspect_ratio override the inherited image-1 aspect ratio.

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+resource ('Create an image from several reference images ... and an instruction'), which clearly separates it from the singular sibling edit_image by emphasizing multiple ordered references. It stops short of naming or contrasting any sibling explicitly, so an agent must infer the boundary from the 'several reference images' phrasing.

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?

Usage is implied rather than stated: the multi-reference, instruction-driven framing suggests when this beats generate_image or edit_image, but there is no explicit when-to-use/when-not-to-use rule or named alternative. Guidance is adequate but left to inference given seven siblings that overlap in the image-editing space.

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

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

Annotations already declare non-readOnly, non-idempotent, non-destructive, closed-world, so the safety profile is covered and the description builds on it usefully. It adds genuinely non-schema behavior: runs fully offline, typical runtime of 45 s to 4 min, and returns the absolute path of the saved PNG. It could have noted whether the model must be loaded or whether generation can be cancelled/queued.

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 short sentences, each load-bearing: purpose, dimension defaults, return value, environment/latency. The core action and model are front-loaded, and nothing is restated from the schema.

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 no-output-schema generation tool the description covers the essentials an agent needs: what it produces, where the artifact lands, how long it takes, and that dimensions resolve automatically. Minor gap: it does not mention that the local model may need to be ready or how failures surface, despite a get_generator_status sibling.

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 100%, so the baseline is 3, but the description adds inference behavior the schema does not: aspect ratio can be parsed from the prompt text (e.g. '16:9 hero image') and dimensions default automatically. That is real added meaning beyond the per-parameter descriptions.

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 ('Generate an image from a text prompt') and names the backing model (local Qwen-Image 2.1), which differentiates it functionally from the edit/inpaint/upscale siblings that take reference images. It never explicitly contrasts itself with those siblings, so an agent must infer the text-to-image vs image-editing split from the verb alone.

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?

Usage guidance is parameter-centric rather than routing-centric: it explains that dimensions are auto-chosen unless width/height or aspect_ratio are given, which tells the agent how to call it but not when to prefer edit_image, outpaint_image, or upscale_image instead. No prerequisites (e.g., model availability) are stated, though get_generator_status exists as a sibling.

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

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

Annotations already declare readOnlyHint=true, idempotentHint=true, destructiveHint=false, and openWorldHint=false, so the safety profile is covered. The description adds valuable context about what status data is available (server, model, GPU, VRAM, job, queue), but does not disclose return format, latency, or error behavior.

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?

It is a single front-loaded sentence with no filler, listing the status categories efficiently. The structure is a noun-clause fragment rather than a full sentence, but it remains clear and appropriately sized for the tool's scope.

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 carries the burden of explaining return values; it covers the major status categories an agent would need (server, model, GPU/VRAM, active job, queue). It does not detail exact response fields or formatting, but given the tool's low complexity and richness of annotations, the coverage is adequate.

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 tool takes zero parameters, so the baseline is 4 per the rubric. There is no parameter information to add or evaluate, and the empty schema is fully consistent with the description.

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 this tool reports generator status and enumerates the key status components: local server running, model loaded, GPU/VRAM, active job, and queue. It distinguishes itself from the generation/edit siblings by being the only status-query tool, though it does not explicitly name or route away from get_image_info.

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 no explicit when-to-use guidance, prerequisites, or alternatives. It implies a status-checking use through the listed return content, but there is no statement such as 'use before generate_image to verify readiness' or 'use instead of get_image_info when checking server health.'

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

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

Annotations already declare readOnlyHint=true, idempotentHint=true, and destructiveHint=false, so the safety profile is covered. The description does add real value by disclosing the conditional branch: generation metadata (prompt, seed, model, steps) is only returned for Kokon-produced images. It does not mention error behavior for missing or non-image paths.

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?

One compact sentence that front-loads the common fields and qualifies the metadata case at the end. No wasted words, though the fragment style means the action itself is never stated.

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 simple read-only inspection tool with no output schema, the description does convey the shape of the returned data, which is the main need. It still leaves the input parameter and failure modes unaddressed.

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 0% for the single required parameter image_path, and the description offers no clarification of it — no path format, no local-vs-URL distinction, no constraints. With low coverage the description is expected to compensate, and it does not.

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 enumerates the fields it returns (width, height, aspect ratio, file type, generation metadata), which lets an agent infer it reads image metadata, but it never states a verb or the resource explicitly and gives no differentiation from the sibling generation tools. It reads as an output-field list rather than a statement of purpose.

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 no guidance on when to call this versus the sibling tools (generate_image, edit_image, upscale_image, etc.) or what prerequisites apply. The only hint is the conditional 'for images made by Kokon', which describes output variance rather than usage conditions.

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

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

Annotations cover the safety profile (destructiveHint=false, non-idempotent), and the description adds real behavioral context beyond them: that pixels outside the mask are preserved exactly and that the model sees the marked region on a copy of the image. It does not discuss failure modes or how the non-idempotent seed affects reruns, so not a 5.

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?

Two sentences, the core constraint front-loaded in the first clause, and the second sentence adds a genuinely useful guarantee. No filler.

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 an 8-parameter generation tool with no output schema, the description adequately explains the transformation and preservation guarantee, leaving quality/seed/output_path details to the well-covered schema. It omits any note on runtime or result handling, a minor gap.

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 88%, so the schema already carries nearly all parameter meaning. The description restates the mask convention (white = change, black = keep) and the four action modes, which duplicates rather than extends the schema. Baseline 3 is correct.

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 precise verb+resource+scope: 'Change only the masked area of an image' with an immediate inline definition of mask semantics. This distinguishes it cleanly from siblings like edit_image (whole image) and outpaint_image (extends canvas).

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 mask semantics imply when the tool applies (partial, localized edits), but there is no explicit routing guidance versus edit_image, edit_images, or outpaint_image. The agent must infer the boundary from the mask requirement alone.

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

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

Annotations declare the safety profile (non-destructive write, not idempotent), and the description adds meaningful behavior beyond that: canvas, mask, and model-aligned sizes are auto-derived, and the original pixels are preserved exactly. It still omits where the new file lands and what the runtime cost is for the three quality tiers.

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?

Two sentences with no padding; the core operation is front-loaded and the automatic-canvas note follows immediately. Every clause carries information.

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 an 11-parameter generation tool with no output schema, the description covers the mechanic but not the result: whether an image file is written, where it is written by default, or how long the quality tiers take is left to the schema and annotations. Adequate, but an agent cannot fully predict the call outcome.

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 55%; the description maps the expansion to per-side values and pixels/percent, which corresponds to top/left/right/bottom and units. However, several parameters (quality, seed, output_path, include_preview) are documented only in the schema, so the description adds little beyond the coverage baseline.

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 ('Extend an image beyond its borders'), which naturally separates it from inpaint (fill inside) and upscale (resize). It never names a sibling tool, so an agent must infer the boundary rather than being told it.

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 explains how to specify the expansion but not when to reach for this tool instead of edit_image, inpaint_image, or generate_image, and gives no prerequisites or exclusions. Usage is only implied by the operation itself.

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

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

Annotations already establish readOnly=false and destructiveHint=false, so the description's job is to add context, which it does: it discloses that a new file is saved, the input is never overwritten, no new content is hallucinated, and runtime varies by hardware. That is meaningful behavioral detail beyond the annotations, though permissions/prerequisites are not addressed.

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?

Three tight sentences, front-loaded with the core action and default, followed by model/timing context and output-safety guarantees. Every sentence carries information, though the model-name detail is the least essential.

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 usefully explains that a new file is written and where output_path constraints apply, and it covers the preview option's default-false behavior. Minor gaps remain around what the tool result returns when include_preview is false, but overall the definition is complete enough to call correctly.

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 100%, so every parameter is already documented in the schema, including the default-2 scale, the 8192 px maximum, and the never-overwrite output_path behavior. The description largely restates those facts, so it adds little beyond the structured fields; baseline 3 applies.

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 names a precise verb and resource ('Enlarge an image') and immediately bounds the operation with the default 2x factor and the 8192 px ceiling, so an agent knows exactly what the tool produces. Upscaling is inherently distinct from the sibling generate/edit/inpaint/outpaint tools, and the description reinforces this with 'adds no new content'.

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?

Usage is implied ('enlarge an image') and the model/timing context ('seconds on a PC GPU, longer on a Mac') helps set expectations, but there is no explicit when-to-use versus when-not, and no routing to or away from siblings like edit_image or inpaint_image. A minimum-viable level of guidance only.

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