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

Annotations already declare readOnly, idempotent, non-destructive and closed-world, so safety needs no restating. The description instead discloses the estimation fallback chain, the 5% exclusion rule for unestimable materials (bentonite, stains, mixed borates), the no-estimate outcome when that share is exceeded, and that nothing else is assumed. That is genuine behavioral disclosure beyond the annotations; only the return shape remains unstated, which is the main gap given there is no output schema.

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 content is relevant and the opening is front-loaded with the primary capability, but the middle collapses into a dense semicolon-and-parenthesis chain covering density heuristics and Appen's glass factors. Those details are useful yet buried, and the single-paragraph block gives the agent no structural handles. Some trimming and ordering by mode would materially improve it.

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 six-parameter, no-output-schema calculation tool, the description covers the derivation logic, the mutually exclusive input modes, the override precedence, and the edge cases where estimation fails. An agent can decide which inputs to supply and predict roughly what comes back. Nothing essential for correct invocation 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 already 100%, so the baseline is 3, but the description adds real meaning on top: drySg wins over the estimated value, lines are used only to estimate dry materials' SG, and gramsPer100ml is the raw measurement alternative to sg. It also explains the estimation basis (mineral density, 2.6 for commercial feldspars and clays, Appen's factors for frits), which the schema does not. This is meaningful added semantics beyond the field descriptions.

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 states a concrete verb and resource: derive a glaze slurry's specific gravity from grams per 100 mL, or take the SG, then compute water to add or pour off, plus Brongniart's-formula dry glaze per litre. The domain is unmistakably distinct from siblings like analyze_recipe or solve_target. It loses a point only because the run-on phrasing forces the agent to parse three capabilities out of a single sprawling sentence rather than a clean enumeration.

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 does give conditional input guidance — 'The water to add needs nothing else', dry weights need drySg or lines — which effectively tells the agent which mode it is in. However, it never names an alternative sibling or states when to prefer this over solve_target or analyze_recipe, and there are no explicit exclusions. Usage is implied through input dependencies rather than stated.

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