flatmark

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

Annotations already declare readOnlyHint and openWorldHint, but the description adds substantial context beyond them: the 8 MB size cap, supported extensions, PDF plain-text limitations without headings or OCR, returned fields, anonymous auth behavior, and credit cost. No annotation contradiction.

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-loaded with the core action and size limit, then structured details on inputs, outputs, limitations, and alternatives. Every sentence contributes unique operational 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?

Given an output schema exists, the description need not define return structure but still names the markdown and meta outputs. It covers size limits, supported formats, PDF caveats, alternative tool routing, authentication, and cost, leaving no meaningful gap for correct invocation.

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 schema already documents all three parameters. The description adds useful selection semantics by explaining the two input modes: base64 via content_base64 or a download link via file, plus that filename's extension determines the parser.

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: convert a document up to 8 MB to Markdown and return it directly. It clearly distinguishes itself from submit_conversion_job by naming that sibling for scans, tables, and page numbers.

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?

Explicitly covers when to use this tool versus the alternative, including the exact conditions for choosing submit_conversion_job. It also states the anonymous rate-limit availability and the 1-credit cost.

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?

With annotations only covering readOnly/openWorld, the description carries real behavioral weight: it enumerates all four status values, explains the 100 KB inline-vs-result_url branch, the failure error contract, the 7-day retention and expiry reporting, and the API-key requirement. This is exactly the context an agent needs to interpret responses.

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-loaded with purpose, then well-organized status/return/retention details. Slightly redundant at the end, where 'Credits: free.' and 'This tool is not metered.' say nearly the same thing, costing some tightness.

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 a single required parameter, existing output schema, and read-only annotations, nothing an agent needs is missing: status semantics, size-based return branch, failure contract, expiry, auth, and cost are all covered.

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% for the single job_id parameter and its description points back to submit_conversion_job, so the schema already does the work. The prose adds no format or syntax detail beyond that, making 3 the correct 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 precise verb+resource ('Check a queued conversion job you submitted') and immediately distinguishes itself from submit_conversion_job by scoping to an already-submitted job. An agent can pick this tool over its siblings without ambiguity.

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 phrase 'you submitted' and the job_id's provenance ('returned by submit_conversion_job') imply the post-submission polling context clearly. It does not, however, state explicit alternatives or when-not-to-use conditions, e.g. that it should be polled repeatedly until terminal status.

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 flag non-idempotent, open-world, non-destructive, non-read-only; the description goes far beyond that with limits, the ~2 minute duration, API-key requirement, credit cost and refund-on-failure policy, file deletion on completion, 7-day result retention, and webhook signature/HMAC/secret-shown-once/3-retry semantics. This is exactly the kind of behavior an agent must know before issuing a paid, non-idempotent job.

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?

Five short paragraphs are front-loaded with the key constraints and each covers a distinct concern (submission, polling, webhook, lifecycle, cost). Minor redundancy exists where the 25 MB limit and allowed extensions restate the schema, but no sentence is 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?

Despite an output schema existing (so return fields need no explanation), the description still names the returned `job_id`/`status`/`result_url` to anchor the polling loop. Combined with limits, auth, cost, retention and webhook contract, nothing an agent needs to invoke this 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%, so the baseline is 3, and the description adds genuine value by explaining the two mutually exclusive ingestion paths (`content_base64` vs. `file`) and the webhook's signature and retry contract. The 25 MB limit and extension list partially repeat schema text, keeping it just below a 5.

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 opening sentence states a specific verb and resource (queue a document for conversion) with concrete scope limits (25 MB, 200 pages) and the processing features (OCR, tables). It also routes the agent forward to `get_conversion_job`, which cleanly separates this async queueing tool from its polling sibling.

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 a clear operational workflow (submit, then poll `get_conversion_job` until `succeeded`/`failed`, or supply a `webhook_url`) and names the follow-up tool explicitly. It never states when to choose this over the sibling `convert_document`, so the alternative-selection guidance is incomplete, but the usage context itself is unambiguous.

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