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

Goes well beyond the annotations by disclosing the response shape difference (md/txt/html/csv return inline `text`, everything else returns a download link valid only 5 minutes), OCR behavior for scanned PDFs and images, and the billing model (first 1000/month free, then 0.01 € each). These are exactly the operational facts an agent cannot get from readOnlyHint=false or destructiveHint=false.

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?

Front-loads the purpose, then packs the format matrix into one dense but scannable clause, followed by return-format and pricing sentences. The conversion lists are long but each entry carries decision-relevant information; the only slight waste is the somewhat verbose arrow chains.

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 return-value burden itself — and it does, covering the text-vs-download-link split, the 5-minute link expiry, and OCR handling. Combined with the format matrix and cost note, an agent has everything needed to call this correctly on the first try.

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, and the description earns a bump by explaining what each `to` value implies for the response type and by making the md choice a usability recommendation. What it doesn't do is clarify format-specific failure modes (e.g., unsupported source/target pairs) beyond the matrix.

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+resource (convert a file) and backs it with an explicit source→target format matrix, so an agent can immediately tell this apart from create_document, upload_file, or get_document. The 'with no AI' qualifier further distinguishes it from AI-driven document creation siblings.

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?

Gives clear routing context: file_id comes 'From upload_file, or the id of a document Doconda made (list_documents)', and 'use md to read any file' steers the agent toward md when the goal is reading content. It stops short of stating when NOT to use it (e.g., when to prefer get_document over converting), so no explicit exclusions.

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 readOnly=false, non-idempotent, non-destructive, so the mutating nature is already covered; the description adds genuinely new behavioral facts beyond that: the call blocks ('Waits until it is ready') and the result is download links. It does not mention failure modes, credential/quota prerequisites, or that repeated calls produce duplicates (implied only by idempotentHint=false in annotations).

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 sentences, front-loaded with the capability and output formats, then the input fork, then the wait/return. Zero filler, though the parenthetical artifact explanation is slightly dense and the prompt/markdown fork could be tighter. Well within size for a 10-parameter tool.

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?

No output schema exists, and the description compensates by telling the agent the call is synchronous and yields download links. Combined with exhaustive schema docs for cost, retention and quality, an agent has enough to invoke it correctly; only error/edge behavior and the create-vs-edit boundary are left implicit.

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 schema already documents all 10 parameters, including enums, defaults, the format→output mapping, the retention/balance-hold nuance and the max_quality/quality relationship. The description only restates the format and prompt/markdown choices, adding no syntax or edge-case meaning the agent could not read in 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 specific verb and resource ('Create a finished document') and enumerates the concrete outputs (Word, PDF, PowerPoint, Excel, artifact), including a plain-language definition of the non-obvious 'artifact' format. An agent can distinguish this from sibling upload_file (which supplies inputs) and edit_document (which modifies an existing one) without opening either schema.

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?

Gives explicit selection guidance: 'Choose only the format' and the prompt-vs-markdown fork ('Give prompt to have it written, or markdown with the content already written'). It stops short of naming edit_document or review_document as the alternatives for non-creation work, so the when-not-to-use boundary is only implied by the verb 'create'.

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 that this is a non-readonly, non-destructive, non-idempotent operation. The description adds useful behavioral context: only requested changes are applied, everything else stays identical, a new file is returned, and get_report can list before/after changes.

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 three compact sentences with the action front-loaded. Every sentence adds either scope, behavioral assurance, or return-value guidance without 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?

There is no output schema, so it is valuable that the description states a new file is returned and points to get_report for the change log. Combined with the rich parameter schema and annotations, it is nearly complete, though explicit sibling routing is still missing.

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%, and the schema already documents file_id, instructions, quality, and max_quality in detail. The description only reinforces that instructions are natural-language requests; it adds no new semantics for the quality controls or parameter constraints.

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: change an uploaded Word, PowerPoint, Excel or PDF file. It includes a concrete example and makes clear this edits existing content rather than creating or uploading a new file.

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 'change an uploaded ... file as asked', but the description does not explicitly say when to choose this over create_document, review_document, or upload_file. The get_report reference explains follow-up inspection, not tool selection.

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 and idempotentHint, so the safety profile is covered. The description adds genuinely non-obvious behavior: the response contains a status and download links that expire after 5 minutes, which tells the agent to fetch and use links promptly rather than cache them. It stops short of describing status values or error behavior for a missing/invalid document_id.

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 short clause with zero waste, and the time-sensitive caveat is included inline. It is a sentence fragment rather than a structured statement of purpose, which costs it a point.

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 one-parameter read tool with no output schema, the description does mention what is returned, which is the main gap it needed to fill. However, it leaves the shape of 'status' and the nature of the download links vague, which is thin for a tool whose whole value is that return payload.

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% with the single document_id documented as 'doc_…', so the schema already carries the parameter semantics. The description adds nothing about the ID format or lookup failure modes, so the 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 the resource and what comes back (status plus download links) but never states the action with a verb, and it does not distinguish this tool from siblings like get_report, read_artifact_data, or list_documents. The title 'Get document' supplies the verb, so an agent can infer retrieval, but the description itself only implies 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?

There is no guidance on when to call this versus get_report, list_documents, or read_artifact_data, and no mention of prerequisites such as needing a prior create_document/upload_file. The 5-minute expiry note is a timing caveat, not usage guidance.

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?

Beyond the annotations, it discloses the return content shape (added/changed/removed passages with before/after text) and important lifecycle states: report_not_ready before completion and content_deleted after retention. It does not cover invalid-ID behavior or the full response structure, but with annotations already declaring read-only/idempotent behavior this is a useful addition.

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 compact and front-loads the core purpose, then follows with the alternative tool and lifecycle states. Every sentence adds useful information without repetition.

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 one-parameter read tool with no output schema, it sufficiently explains the main returned content and special report states. It could more fully specify the changes object or edge cases, but the description covers what an agent needs to invoke and interpret it 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?

The schema has 100% description coverage for the single document_id parameter, so the schema already documents it. The description does not add any further parameter-level meaning, making the baseline score 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 concretely defines the tool as retrieving the change report produced by edit_document, including what the changes list contains. It also explicitly distinguishes it from get_document, which is for status and download links.

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?

It names the alternative get_document and the condition for using it (status and download links), which gives clear routing guidance. It implies the use case for get_report through the first sentence but does not spell out a full when-to-use/when-not-to-use rule.

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 readOnly, idempotent, and closed-world, so the safety profile is covered. The description adds genuinely new context: results are project-scoped, sorted newest first, and carry id and name. No pagination or exhaustion behavior is mentioned, which is the only real gap.

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 tight sentences, no filler. The core output (id + name, newest first) is front-loaded and the follow-up sentence covers downstream use.

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?

Single optional parameter, no output schema, and the description itself describes the return shape (id and name) plus ordering and scope. Nothing an agent needs to invoke it correctly is missing.

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% and the limit parameter already documents range and default. The description only implies ordering is fixed ('newest first'), which is a small semantic addition beyond the schema — baseline 3 fits.

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 plus scope and ordering: 'latest documents made in this project (newest first), with their id and name.' An agent immediately knows what comes back. It doesn't explicitly contrast with get_document, but the 'latest/made in this project' framing distinguishes it from a single-document fetch.

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?

Gives concrete downstream routing: 'Use an id as file_id to convert or edit one, or in source_file_ids to create from it,' which points at convert_file, edit_document, and create_document. It stops short of stating when to pick this over get_document when the id is already known.

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 openWorldHint=false, so the safety profile is covered. The description adds real value beyond that by disclosing the result shape (collections vs. records) and the 'newest first' ordering, but says nothing about pagination, result limits, or what an empty artifact returns — meaningful gaps for a records-listing read.

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?

A single well-formed sentence with no filler, and the core resource is front-loaded before the returned shapes. It is slightly compressed to the point of being terse ('What people have saved in an artifact's page'), but nothing is wasted.

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 only two parameters, full schema coverage, and no output schema, the description supplies the essential missing piece — what the call actually returns (collections or records, newest first). Remaining gaps (pagination, volume, permissions) are minor for a simple two-parameter read tool whose annotations already signal a safe, idempotent operation.

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 both parameters are already documented in structured data, establishing a baseline of 3. The description's mention of 'one collection's records' largely restates the schema's own 'Optional: one collection's records instead of the list of collections' and adds no naming format, matching rule, or error behavior.

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 concrete resource — what users have saved on an artifact's page, such as poll votes — and states the two shapes it can return (a list of collections, or one collection's records). The retrieval verb is only implied by the name/title rather than stated in the body, and no sibling is named, but the resource is specific enough to separate it from get_document or get_report.

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 the phrase 'what people have saved in an artifact's page,' which tells an agent this is the tool for reading submitted/collected data rather than document content. There is no explicit when-not guidance, no mention of prerequisites (does the caller need access to the artifact?), and no alternative tool named for overlapping cases like get_document.

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?

The annotations already declare this as a non-read-only, open-world, non-idempotent, non-destructive operation. The description adds useful behavioral context beyond those hints: accepted file types, a 20 MB size cap, and the fact that it returns a file_id. It does not mention auth or rate limits, but with annotation coverage this is a solid disclosure level.

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 front-loaded with the action and key constraints, and every sentence carries useful information. The long parenthetical format list is somewhat cumbersome but justified for an upload tool, so it is concise enough without being exemplary.

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 appropriately fills the return-value gap by stating that it returns the file_id. It also covers formats and size limits, which are important for correct invocation; the main remaining gap is that it does not explain how to choose among url, path, and content_base64, though the schema covers those fields.

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 all four parameters are already documented in the schema. The description only adds a small path hint that largely repeats the schema's "Absolute path of the file on this computer" text and does not clarify mutual exclusivity among url, path, and content_base64; baseline 3 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 gives a clear verb and resource ("Upload a file") and adds concrete scope: supported formats, the 20 MB limit, intended downstream uses, and the returned file_id. It is specific enough to distinguish this as the file-ingestion tool, but it does not explicitly contrast itself with siblings such as convert_file or create_document.

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?

It states clear use contexts: upload to convert or edit the file, or to use it as a source. However, it does not name alternatives or exclusions, such as when to use upload_file versus convert_file or create_document, so it falls short of full routing guidance.

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