Compare models

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

With no annotations, the description carries the full burden and largely meets it: it discloses how missing cells are footnoted so no gap is silent, that arena scores are not cross-comparable, that mixing model types is refused, that refresh bypasses the cache, and that a reproducible spec block is always returned. It does not state auth needs or rate limits, leaving a small gap.

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 purpose and scope are front-loaded and the Args section is well organized, but the opening prose paragraph is long and dense with operational detail that arguably belongs in contextual notes rather than a purpose statement. Most sentences carry information, yet the description leans toward over-specification relative to a concise tool blurb.

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?

With no output schema and no annotations, the description covers what an agent needs: input ranges, view modes and their rendered output (columns vs. per-metric deltas), the default metric set, arena behavior, refusal conditions, caching, and the guaranteed spec block. Nothing essential for a correct call 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 description coverage is 0%, so the description must compensate and it does: each of the six parameters is documented with accepted values and defaults (metrics accepts 'all'/'benchmarks'/comma lists, view is 'table'/'diff', compact is ignored when metrics is given, spec merges with explicit args, refresh bypasses cache). It adds semantics well beyond the bare schema types.

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 ('Side-by-side comparison of 2-12 models on any benchmark'), names the input domain (language models or arena entries), and mirrors a concrete external artifact (artificialanalysis.ai comparison pages). An agent can distinguish it from aa_leaderboard, aa_model, and aa_pareto 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?

Usage is implied by the compare semantics and edge cases are described (arena-only comparisons render board metrics; mixing a language model with an arena entry is refused with an explanation). However, there is no explicit routing guidance telling the agent when to pick aa_compare over aa_leaderboard, aa_pareto, or aa_model, which is the core 'vs alternatives' question.

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