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

Annotations already provide readOnlyHint, openWorldHint, and idempotentHint, so the bar is lower. The description adds substantial behavioral context: it discloses that null constraint fields are reported as such and never fabricated, that v4 constraint is flagged beta, that constraint_flags surfaces caveats, and that the effective dataset and reference build are echoed. These details address output edge cases and data-source caveats beyond what annotations convey, without contradicting them.

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 description is information-dense and front-loaded with the core purpose, then systematically elaborates on metrics, version differences, null behavior, and data source. Every sentence serves a purpose, with no filler. It is somewhat long but justified by the tool's complexity. A 4 reflects strong structure with minor verbosity, not a flaw.

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?

The output schema exists, so return-value explanation is not required. The description covers the core metrics, version-specific thresholds (pLI >0.9, LOEUF <0.6 v4 / <0.35 v2), null-field handling, beta status, covariance with reference genome, and the data source attribution. There are no obvious gaps that would prevent an agent from correctly invoking and interpreting the tool.

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 baseline is 3. The description reiterates the accepted gene identifier types (HGNC symbol or Ensembl ID) and mentions echoing dataset/reference build, but it does not add new meaning beyond the schema for parameters. The threshold values (LOEUF <0.6 in v4) are interpretation aids, not parameter semantics. Credit is given for reinforcing schema details, but no substantive additions.

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: 'Fetch gnomAD loss-of-function constraint for a gene' and enumerates the exact metrics returned (pLI, LOEUF, observed/expected ratios, Z-scores). It explicitly differentiates itself from the allele-frequency axis, making clear it is the orthogonal tool for gene tolerance. This distinguishes it from sibling tools like gnomad_get_variant and gnomad_list_gene_variants 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 description provides strong contextual guidance on when this tool is relevant ('a loss-of-function variant matters far more in a gene intolerant to being broken') and notes the beta flag for v4 constraint. However, it does not explicitly name alternative tools or state when NOT to use this tool. The schema also covers required input types. This is clear context but lacks explicit exclusions or sibling routing, so a 4 is appropriate.

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