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

With no annotations, the description carries the full burden and does: it discloses that counts reset on process restart, that reading is non-mutating ('reading them never changes them'), and that the two series are never merged — all behavior an agent could not infer from the empty schema.

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 what is reported before scope and caveats, and every clause adds information (separation rule, restart behavior, read-only nature). The single dense block is slightly long, but no sentence is filler.

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?

No output schema exists, and the description supplies the return shape (hit/miss counts, total lookups, hit rate, per-tool breakdown), so an agent knows what it will get without needing a 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?

Zero parameters, so the baseline is 4; there is nothing for the description to clarify, and it does not fabricate parameters.

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 and resource ('Report ... hit/miss counts, total lookups, and hit rate') plus the exact scope (the ~5-minute recommendation cache) and the split granularity (recommend_model vs POST /router/recommend, never combined). An agent can distinguish it cleanly from pricing and telemetry siblings.

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?

Explicitly frames the use case ('confirming the cache is reducing recomputation in production') and excludes a whole class of alternatives ('not a pricing question'), which steers agents away from compare_costs/estimate_cost. It stops short of naming specific alternative tools for other cache/inspection needs, so it is clear context rather than a full when/when-not map.

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