Pareto trade-off frontier

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

With no annotations the description carries the full burden and does well: it defines the dominance rule, the ordering (x ascending), the default cheapest-first truncation, that deprecated rows are excluded by default, and that refresh bypasses the cache. It does not state freshness/rate-limit behavior or the exact response shape, but the cached-vs-refreshed distinction and default filtering are real disclosures.

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 core definition and dominance rule, then organized into site-view mapping and Args sections so an agent can scan it. The site-view block is slightly verbose for three examples and the smart-quote formatting is noisy, but no sentence is purely 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 does explain what is returned (non-dominated models, x-ascending order, knee line, coverage) as well as default truncation and filtering. It stops short of describing the concrete response structure or error behavior, leaving a small gap for a 10-parameter analytical tool.

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% and all 10 parameters are undocumented in the schema, yet the Args block defines every one, including enumerated values for x (cost_per_task, index_total_cost, out_tokens_per_task, time_per_task, context_tokens, params_b, ttft, price_in, price_out) and example values for score. This is exactly the compensation the zero-coverage schema requires.

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/resource (compute the Pareto frontier of any score against any trade-off axis) and quantifies the output (frontier points plus a knee and coverage). This is clearly distinguishable from aa_leaderboard or aa_compare, which rank rather than compute a non-dominated set.

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 tells the agent it is 'Always available — no per-benchmark tool needed' and maps three concrete site views onto parameter combinations, which is strong when-to-use context. It never excludes or routes against specific siblings such as aa_compare, so it stops short of full 5-level routing guidance.

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