Server Quality Checklist

Tool Scores

• generate_image_openaiA
  4.1/5.0
  Behavior3/5

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

  With no annotations provided, the description carries the behavioral disclosure burden. It discloses key behaviors: only absolute output directories are accepted, relative paths rejected, the file naming pattern, and writes confined to the absolute_output_directory. However, it doesn't describe what happens on failure, whether generations are synchronous, API cost implications, or the return value/format. For a generation tool handling disk writes and API calls, some richer operational detail would help.

  Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

  Conciseness4/5

  Is the description appropriately sized, front-loaded, and free of redundancy?

  The description is well-organized with a clear summary paragraph followed by a bulleted parameter guide. The front-loaded purpose sentence is strong. The parameter notes are genuinely useful (size constraints, quality enums per model, background constraints). It's slightly verbose but every sentence earns its place, given the complex cross-parameter constraints.

  Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

  Completeness4/5

  Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

  This is a complex tool: 7 params, model-specific constraints, disk-writing behavior, and nested images array. The description and rich schema (100% coverage) together cover the input side well. No output schema exists, so the description should note return behavior, which it does partially ('written to output directory', file naming). It doesn't state whether it returns success status, generated URLs, or file paths, which is a minor gap for API-integration workflows.

  Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

  Parameters3/5

  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 even with minimal description-level parameter info. The description does add helpful cross-parameter constraints (e.g., transparent background requires png/webp, gpt-image-2 doesn't support transparent, quality varies by model family) that go beyond individual parameter schemas. These relational constraints add value, though the schema already documents most per-parameter details thoroughly.

  Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

  Purpose5/5

  Does the description clearly state what the tool does and how it differs from similar tools?

  The description opens with a specific verb+resource: 'Generate one or more images with OpenAI's Image API and save them to disk.' It clearly states the tool's function and scope (website/app assets), and distinguishes it from the sibling generate_image_gemini by explicitly naming 'OpenAI's Image API'. The model list in the description further differentiates it from the Gemini sibling.

  Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

  Usage Guidelines5/5

  Does the description explain when to use this tool, when not to, or what alternatives exist?

  The description gives explicit usage context: 'Use this when the user (or your plan) needs OpenAI-generated assets for a website or app.' This clearly signals the when-to-use scenario and implicitly distinguishes from the Gemini alternative. While it doesn't name generate_image_gemini explicitly as an exclusion, the OpenAI-specific framing makes the selection obvious.

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

• generate_image_geminiA
  4.3/5.0
  Behavior4/5

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

  No annotations are provided, so the description carries the behavioral disclosure burden. It effectively discloses key behaviors: images are written only to the specified absolute_output_directory, relative paths are rejected, the file naming convention ({filename}-gemini-{model}-{timestamp}.{ext}) is documented, and the mime_type constraint (only image/jpeg) is surfaced. It does not disclose potential side effects like file overwrites or API cost implications, but the core behavioral contract is well-covered.

  Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

  Conciseness4/5

  Is the description appropriately sized, front-loaded, and free of redundancy?

  The description is well-organized with a clear opening sentence followed by a requirements list and a notes section for format options. It front-loads the purpose and requirements. The notes on aspect_ratio, image_size, and mime_type are useful but partly redundant with the schema enums, though they add the uppercase-K requirement which the schema also mentions. This is a compact, scannable structure, though the inline enumeration is slightly repetitive with the schema.

  Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

  Completeness4/5

  Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

  For a tool with 6 parameters (3 optional with enums), no output schema, and no annotations, the description does substantial work. It documents the file naming convention, the absolute-path requirement, the model variants, and all format constraints. The main gap is that it doesn't describe the return value/behavior after generation (e.g., does it return paths, a count, errors?), which could matter for an agent chaining subsequent steps. Overall this is quite complete for the tool's complexity.

  Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

  Parameters4/5

  Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

  Schema coverage is 100%, so the baseline is 3. The description adds value beyond the schema by enumerating all allowed aspect ratios inline, clarifying the uppercase-K requirement for image_size (which is a common failure point), and explicitly stating that only image/jpeg is supported. It also documents the filename suffix convention which maps to the 'filename' parameter semantics, adding meaning beyond the schema's 'Base filename without extension.'

  Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

  Purpose5/5

  Does the description clearly state what the tool does and how it differs from similar tools?

  The description clearly states the verb+resource ('Generate one or more images with Google Gemini (Nano Banana) via the Interactions API and save them to disk'). It explicitly distinguishes from the sibling tool by naming the provider (Google Gemini vs OpenAI) and the model family (Nano Banana), which separates it from generate_image_openai. The mention of 'Interactions API' and 'Nano Banana' adds specific identifying context.

  Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

  Usage Guidelines4/5

  Does the description explain when to use this tool, when not to, or what alternatives exist?

  The first sentence provides a clear use case: 'when the user (or your plan) needs Gemini-generated assets for a website or app.' This establishes when to use it. It doesn't explicitly exclude OpenAI/alternatives, though the image model id examples and 'Gemini' naming imply the sibling tool handles other providers. It clearly states requirements (model id, prompt+filename pairs, required absolute_output_directory) which helps the agent assemble correct calls.

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