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

Annotations already cover the safety profile (readOnly, idempotent, non-destructive, openWorld), so the description is free to add substance — and it does: data provenance (PubTator 3 BioC JSON export at NCBI/NLM), the caveat that all annotations are machine-extracted, that no API key is required, the 20-PMID cap, and that full text is PMC open-access only and truncated to 1,500 chars per passage. These are exactly the traits an agent cannot read off the annotations.

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 purpose, then layers the return contents, the size cap, the source, and the keyless note. Dense but every clause carries information; the only minor cost is that the middle enumeration is long and slightly list-heavy.

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 carries the return-shape burden and does so well — it names the entity types, the normalization targets, character offsets, and the presence of relations. Minor gaps remain: behavior on invalid/unknown PMIDs and how relations are structured in the payload are unspecified, but the essentials for correct invocation are present.

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 both parameters, including the 1-20 PMID bound and the full-text truncation behavior. The description largely restates those facts ('Up to 20 PMIDs per call', PMC full text) and adds only that full text replaces title+abstract, so the 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+resource — entity annotations for specified PubMed articles — and enumerates exactly what is returned (gene/disease/chemical/variant/species/cell-line mentions, normalized ids, offsets, extracted relations). The PMID-driven framing cleanly separates it from keyword-oriented siblings like pubtator_search and relation-only siblings like pubtator_relations. An agent can identify the tool without opening the 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?

Usage is implied rather than stated: 'for specific PubMed articles' signals this is the path when PMIDs are already known, and 'Up to 20 PMIDs per call' bounds the batch. However, it never names an alternative (pubtator_search, pubtator_relations, pubtator_find_entity) or states when not to use it, leaving routing to inference.

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