find_overlaps

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

Annotations already declare readOnly, idempotent, and non-destructive, so safety is covered; the description adds genuinely new context by framing the output as a ranking requiring verification rather than an authoritative answer. The distinction between binding and out-of-force clauses, and the instruction to handle each differently, is behavioral guidance not present in any structured field. It does not, however, say what the ranking is based on or how confident the matches are.

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?

Compact and front-loaded: the trigger condition ('Before you write a decision') comes first, followed by input instructions, then interpretation guidance. The parentheticals about `treats` and replacement citations are dense but each carries actionable information, so no sentence is wasted.

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 three-parameter read tool with full annotation coverage and an output schema, the description does the necessary work: trigger, inputs, and how to interpret the result. The only gaps are the unexplained `limit` parameter and the absence of any statement about ranking criteria, which the output schema may partially offset.

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 0%, so the description must carry all parameter meaning. It clarifies `draft` ('paste the draft text') and `paths` ('the files it touches'), including that either may be used, but never mentions `limit` or how the two inputs interact when both are supplied. Two of three parameters get loose treatment, leaving meaningful gaps.

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 asks 'which existing clauses does your draft overlap?' and specifies what gets flagged (clauses that still bind vs. clauses out of force), which pins down the verb and resource well enough that find_overlaps is unambiguous. It also separates itself from propose_decision by labeling its output 'a ranking, not a ruling,' though the purpose is delivered as a rhetorical question rather than a direct statement.

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?

It places the tool precisely in the workflow ('Before you write a decision') and tells the agent what to feed it ('Paste the draft text and/or the files it touches'). It also names the follow-up step ('then verify with propose_decision') and gives downstream actions for each flag type (declare in `treats`, cite the replacement). It stops short of stating when not to use it, e.g., a draft with no existing related clauses.

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