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

Annotations declare idempotent=false and readOnly=false. Description adds critical behavioral context: cost (5 credits), exact return value (task_id), polling requirement, and specific resultType parameter needed for fetching. Does not contradict annotations.

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?

Three sentences dense with essential information. Front-loaded with purpose, followed by workflow, then constraint. Slightly dense but no waste given the complexity of the async lifecycle that must be communicated.

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?

Despite no output schema, fully compensates by documenting the task_id return value and complete polling workflow (status check → result fetch). For a single-parameter async task starter, this is comprehensive.

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 has 100% coverage (baseline 3). Description adds semantic value by specifying the model_id must be from an 'existing' custom eval model (constraint not explicit in schema) and links model_id to the resource_id parameter in the result fetching workflow.

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 specific action 'Start async inference', target resource 'custom evaluation model', and outcome 'auto-generate evaluation dimensions'. Also notes cost '5 credits'. Clearly distinguishes from sibling tools like atlas_start_custom_eval_batch by focusing on single-model inference.

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?

Provides explicit async workflow guidance (poll with careerproof_task_status, fetch with careerproof_task_result) and prerequisites (model_id must be from existing model). However, lacks explicit comparison to sibling atlas_start_custom_eval_batch for when to use batch vs single inference.

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