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

Annotations only state openWorldHint/non-idempotent/non-destructive; the description adds genuinely non-obvious behavior: the registry does the matching, the customer prepays at checkout_url and 'no money moves inside this tool', and check mode judges answers blind (A/B/C) with a four-field signed letter verifiable at /verify. This is exactly the kind of economic and workflow disclosure annotations cannot carry.

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 return values, then the check-mode variant is clearly flagged with 'CHECK MODE:'. It is dense (emoji-like separators, mid-sentence tree structure) but nearly every clause carries routing or return-value information; the answers=[{model, text}] restatement is the only part that overlaps 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?

With no output schema, the description correctly covers the return surface (price, checkout_url, badge_url, letter_url) and the two operating modes. For a 13-parameter, externally-fulfilled job, it omits what happens after prepay (reviewer identity/timing), how max_price_usd and deadline_minutes shape the outcome, and whether a job can expire or be cancelled — modest but real gaps.

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 46%, so the description must do work, and it does: brief is defined as what to check/for whom (and as the literal question in check mode), answers is explained as 1–3 model answers judged blind, and the kind enum values are restated with license implications. It leaves max_price_usd, deadline_minutes, language, and ai_vendor/customer_type semantics to the schema (the latter two have no schema 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 states a specific verb+resource (request a review of an AI output from a registry-verified licensed human) and spells out the resulting scope: registry picks the card, opens the job, returns price, checkout_url, and badge_url. The opening phrase 'One call, no search' implicitly routes an agent away from search-style siblings (search_human, search_cases), and the licensed-credential/sign-off angle separates it from ask_human/call_human.

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 gives an explicit trigger: 'Use this when a policy or a regulator requires a qualified human in the loop for a high-stakes output', and the CHECK MODE paragraph adds a clear second use case (blind judging of 1–3 model answers). It doesn't name alternative tools or state when-not to use it, so it stops 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.