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

Annotations already establish readOnly/idempotent/no-open-world behavior, and the description adds real context: the return is a calculation receipt containing inputs, results, assumptions, validation checks and source, and it is explicitly unsigned and not independent certification. It also discloses the absence of private data access, external lookup and arbitrary code execution. What it does not disclose is any failure behavior for malformed probability rows or non-summing likelihoods.

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 core capability is front-loaded, which is good, but the paragraph is dense and mixes capability description with input-formatting rules and redundant disclaimers. "The receipt is unsigned, not independent certification" largely restates the preceding sentence, and "optional imperfect-signal information value" is a hard phrase to parse on first read.

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 an output schema present, the description needn't document return values, and it does supply behavioral and formatting context appropriate to a read-only calculator. However, for a tool whose correctness hinges on choosing between two mutually exclusive input modes, the description never mentions the modes at all, leaving the agent to discover them from the schema's oneOf. That is the main completeness gap.

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?

Top-level schema description coverage is 0%, so the description must carry the load, and it does explain that tool fields belong inside the nested "inputs" object and warns against passing a REST-only field. That structural guidance is genuinely useful, but it does not explain the meaning or selection of the two `mode` values (perfect vs sample) or when `signalLikelihoods` becomes required, which is the highest-risk parameter decision. The conditional "units or modelVersion only when listed in inputSchema" phrasing is also ambiguous given `units` is an empty object.

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?

"Calculate expected values, best current expected-value action, perfect-information upper bound... and net value after information cost" names a specific verb and a precise set of computed outputs, so the agent knows exactly what comes back. It does not name or distinguish itself from plausible siblings such as buscalc_uncertainty, buscalc_sensitivity or buscalc_probability, which is what keeps it from a 5.

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 clear scoping exclusions ("does not infer probabilities, certify decisions or model risk aversion") and states it handles both maximize-payoff and minimize-cost framings, which tells the agent when the tool fits. It stops short of naming an alternative tool for the excluded cases, so the routing guidance is incomplete rather than explicit.

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