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

readOnlyHint already declares it a safe read, and the description still adds real behavioral detail: filters combine and persist across pagination, counts reflect stored_comparability_records rather than a live inference inventory, and extra row fields (version_count, source_count, first_edition_seen, last_edition_seen) appear only for multi-form/multi-source records. The prose gets tangled around 'FORM's wording source' and doesn't state ordering or result-size behavior, so it stops short of 5.

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?

Purpose and the get_comparability handoff are front-loaded, which is good, but the middle section explaining conditional row fields is dense, repetitive ('reads exactly as it always has') and spends several sentences on one point. It is informative rather than padded, but it is not tight.

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-value burden and largely meets it: it explains counts vs. records, the empty-result answer, conditional presence of version/source-count fields, and that full forms come from the one-record get. Remaining gaps are ordering and default page size handling, the latter covered in the schema.

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 100%, so all 13 filters are already documented with examples and edge notes (e.g. sending_organization_id covering sources with no IPEDS UNITID). The description adds one genuine semantic beyond that — filters combine and remain in effect across pagination — but otherwise re-lists the facets already in the schema. Baseline 3 is right.

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 pair (count and list) on a specific resource (stored comparability bundles) and enumerates the filterable facets: type, sending/receiving institution, source authority, known/inferred kind, agreement, learning unit, exam. It also explicitly separates itself from get_comparability ('Read full evidence with get_comparability'), so an agent can route between the list and detail views without opening either 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?

Names the alternative (get_comparability) and the condition that selects it, and scopes the tool as operating on stored records rather than triggering live four-source inference. It also tells the agent how to read an empty result ('Answers none held when the selected stored records are absent'). No negative guidance on when this search is the wrong entry point, which keeps it at 4.

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