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

Annotations only declare readOnlyHint=false, openWorldHint=true, and idempotentHint=false. The description goes well beyond: it explains that attribution is tied to the key, not declared fields; that submissions are stored as 'pending' with AIO review before publication; that published runs are labelled 'agent-submitted' and never merged with curated results; and that publication is self-reported, not certification. This rich behavioral context is vital for an action with side effects.

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 a compact paragraph but each sentence carries weight: purpose, prerequisite and attribution, completeness rule, processing and review, publication semantics, disclaimer, and a user-consent directive. It is front-loaded with the purpose and maintains a logical flow from invocation to post-submission behavior. No fluff.

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?

Given the tool's complexity (multi-layer submissions, key-based attribution, async review, publication semantics) and no output schema, the description covers all critical aspects: prerequisites, constraints, the review pipeline, the 'agent-submitted' label, the disclaimers, and the ask-user directive. The only minor omission is the exact response format, but the description sufficiently explains what happens to the submission.

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%, so the baseline is 3. The description adds meaning beyond the schema by clarifying the layer-completeness constraint (all 105 items in a layer must be present or the layer omitted) and the attribution semantics of apiKey (run attributed to the key's holder, not declared values). This supplements the schema's structural descriptions, though it doesn't detail each parameter's format beyond that.

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+resource: 'Submit answers to the agent-track item set from get_bench_items.' This clearly distinguishes it from sibling tools like submit_eval by naming the exact source item set. The title echoes and complements this, leaving no ambiguity about the tool's function.

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 states the prerequisite (AIO agent key with bench:submit scope), references the source data (get_bench_items), and gives an explicit instruction to 'Ask the user before calling this.' It also specifies the layer-completeness rule. However, it does not explicitly contrast with alternatives like submit_eval, so the 'when-to-use vs alternatives' guidance is implicit rather than explicit.

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