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

Annotations already cover read-only/idempotent safety, and the description adds substantial behavior beyond them: known false positives, a read guard (version 2) that withholds ten false shapes and reports them in declined_by_read_guard/read_guard_note, hand-read precision with confidence intervals, unknown recall, and the fact that only five labels are loaded. It also discloses that the overruling opinion itself is usually absent (Roe has no Dobbs row), which is exactly the kind of non-obvious behavior an agent needs.

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 essential definition and the id caveat, but the block is extremely dense and carries precision/recall confidence intervals and stratum-specific figures that matter more for trust calibration than for invoking the tool. Several clauses (sample sizes, intervals) could be moved to the referenced coverage statement without losing invocation-critical information.

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?

There is no output schema, yet the description documents the return shape in detail: every row carries the citing court's sentence, by_label reports all five labels regardless of filter, also_carried_by, unrelayable_in_window, capped, plus the mandatory 'note' and coverage_note, and ordering by citing-opinion date. Given the tool's complexity and high false-positive risk, nothing an agent needs to interpret results 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 100%, so the schema already documents all four parameters, including the label enum note 'A filter over sentences, never a verdict'. The description reinforces the label-as-filter framing and the cluster-id-not-opinion-id distinction, but largely restates schema content rather than adding new parameter syntax or semantics; 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?

Opens with a specific verb+resource (later courts reporting a case was overruled/abrogated/etc.) and immediately disambiguates the identifier: 'Takes a cluster id (the case-record id), not an opinion id.' It also distinguishes itself from the sibling concept citation/Shepardizing, so an agent can place it against citation, citations and brief_check without opening a 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?

Explicit exclusions are given ('Not a validity verdict and not Shepardizing', 'zero rows is not evidence that a case remains good law', and the caveat that ~68.7% of findings are expressly partial), plus a temporal boundary explaining thin recent results. It stops short of naming a specific sibling to use instead when you want a positive/validity signal — it only delegates that to 'sources' for coverage.

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