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

Annotations already establish readOnlyHint=true and openWorldHint=true, so the safety profile is covered; the description adds substantial context beyond that: exactness per model family, a hard input ceiling ('Up to 2 million characters'), chat-message overhead accounting, and an explicit per-call price with the x402 payment scheme. Price and input limits are exactly the operational facts an agent needs before invoking.

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 action, then layered with model coverage, limits, and cost in a compact block with no filler sentences. It is dense — a single colon-and-slash-laden run-on — but every clause carries distinct decision-relevant information, so nothing 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?

With no output schema, the description still covers the key pre-call facts: accuracy per model family, input size limit, and cost. The remaining gap is the return shape (a single count versus a per-model breakdown) and whether text and messages are mutually exclusive, but for a two-parameter utility the definition is otherwise complete.

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 50%: 'messages' carries a description while 'text' is bare. The description partially compensates by referencing chat-message overhead and the 2M-character ceiling, indirectly clarifying the text-vs-messages distinction, but it never defines the two parameters or says whether both may be supplied together. With names self-describing and half the schema documented, baseline 3 is appropriate.

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?

Starts with a specific verb+resource ('Count tokens') and immediately scopes it with model-specific detail (o200k for gpt-4o/4.1/o-series/gpt-5, cl100k for gpt-4/3.5, estimates for Claude/Llama/Gemini). This is unmistakably distinct from siblings like chat_completions, embed, or embeddings, which generate or vectorize rather than count.

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 phrase 'before you call a model' gives a clear triggering context: pre-flight cost/context budgeting ahead of a chat or completion call. However, it names no alternatives and states no exclusions (e.g., when to prefer a provider's own tokenizer or when the estimate is unsafe), so guidance is contextual but not comparative.

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