Velrim

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

Goes well beyond the annotations (readOnlyHint=false, openWorldHint=true, idempotentHint=false) by disclosing the return contract: a typed object with per-field state, calibrated confidence, and source anchors. The calibration claim is even backed by a reference URL, so the 0.9 semantics are defined rather than asserted.

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 front-loaded sentences: what it does, what it returns, and how to use the return. Every sentence carries distinct information and none restates the name or 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?

An output schema exists, yet the description still supplies the non-obvious behavioral facts an agent needs (confidence calibration, per-field state, anchors). Input modes, required schema, and the escalation workflow are all covered, leaving no material gap for calling 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?

Schema description coverage is 100%, so the schema already documents all five parameters including the required 'schema' object and the mutually exclusive base64/upload_key pair. The description reinforces the document-input mechanism but adds no syntax or format detail beyond the schema, so 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?

States a specific verb and resource ('Extract structured data from a document') and immediately distinguishes the two supported input modes (PDF/image via document_base64 or upload_key). An agent can tell this apart from the sibling velrim_job_status, which is a polling tool.

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 operational guidance on when to act versus escalate ('act on a field at or above your accept threshold, escalate the fields below it and any that are missing'), plus the input-mode choice. It never names the sibling or an alternative tool or an explicit when-not, 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.

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

Annotations already cover the safety profile (readOnly, idempotent, openWorld), and the description adds substantive behavior beyond them: the polling lifecycle, that the result only appears once the job completes, the per-field result shape (state/confidence/anchor), and how calibration curves should drive accept-vs-escalate branching. This is unusually rich operational context.

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 polling action and the sibling distinction, which is exactly what an agent needs first. The final sentence on reliability curves and branching is longer and advisory, but it still carries actionable decision guidance rather than 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?

An output schema exists, yet the description still summarizes the return contract (state, confidence, anchor) and explains how to act on it. For a single-parameter polling tool, nothing an agent needs to call 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% for the single job_id parameter, so the baseline is 3. The description adds origin and traceability context ('from POST /v1/jobs', 'velrim_extract itself is synchronous and never creates jobs') that helps an agent source a valid value rather than guess.

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 opens with a specific verb and resource ('Poll an asynchronous extraction job by its job_id') and immediately contrasts itself with the sibling tool by noting velrim_extract is synchronous and never creates jobs. An agent can route between the two 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?

It explicitly states the condition that selects this tool (a job_id from POST /v1/jobs, or a job created via the Velrim API/SDK) and names the alternative plus why that alternative is not applicable. The when-to-use boundary is unambiguous.

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