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

Annotations already declare readOnlyHint, idempotentHint, and destructiveHint false, and the description adds that it 'reads stored data only' and is free, reinforcing the read-only safety profile. It also discloses a key behavioral limitation: no result is produced until at least two audits are stored. This goes beyond the annotations by explaining the preconditions and the operational context of the diff.

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 dense and well-structured: core behavior first, then usage timing, then exclusions and cost. It loses a point for minor redundancy — 'call it after re-auditing' and 'USE THIS after a deploy plus a fresh lekta_audit' convey overlapping guidance. Still, every sentence carries operational value, so it remains above average.

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 one-parameter tool with no output schema, the description gives the agent everything needed: what the result contains (grade and score movement, improved/regressed checks), when to call it, when not to call it, and the required precondition (at least two stored audits). Cost and read-only behavior are also stated. Nothing important is missing for correct invocation and interpretation.

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% — the single required param 'url' is documented as 'Absolute http(s) URL with at least two stored audits.' The description adds nuance by emphasizing it must be one URL, not two, and clarifies that the URL is the anchor of the diff operation. This adds semantic framing beyond the bare schema description.

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: 'Compares the two most recent stored audits of one URL' and then specifies exactly what is compared — grade and score movement, which checks improved, which regressed. It also positions the tool relative to its sibling lekta_audit by framing it as the follow-up that proves a fix worked, so an agent can distinguish it from the other lekta tools.

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 is explicit about when to use it: 'call it after re-auditing to prove a fix worked' and 'USE THIS after a deploy plus a fresh lekta_audit.' It also provides concrete exclusions: 'DO NOT use it to compare two different URLs' and warns not to expect a result when only one audit exists. That is clear when-to-use and when-not-to-use guidance.

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