Two tools with basic schemas and descriptions, but significant gaps in parameter documentation and error handling guidance. get_task has optional parameters with minimal context; check_pricing lacks clarity on defaults and return structure. No error recovery hints, no output schema documentation, and descriptions are generic. Tool names follow verb_noun convention but descriptions don't explain WHEN to use each tool or what happens on failure. Parameters lack type constraints (enums, ranges) and dependency documentation.

• Document output schema for both tools. E.g., get_task returns {task_id, status: 'pending'|'completed'|'failed', result: {...}, created_at, updated_at}. check_pricing returns {model, endpoint, price_per_unit, currency, pricing_url}.
• Add enum constraints to 'action' and 'model' parameters. Replace vague descriptions with: 'action: One of the asynchronous endpoints for this model line (e.g., generate_image, upscale). Required if multiple endpoints exist.'
• Expand descriptions to 100-200 chars with WHEN guidance. E.g., 'Call get_task to poll the status of an async image generation task. Returns status (pending/completed/failed) and result URLs when ready. Use after creating a task with an async endpoint.'
• Add error recovery hints. E.g., 'If task_id not found, verify the ID matches the response from task creation. If action is invalid, call list_endpoints() to see available options.'
• Document parameter dependencies. E.g., 'If this model line has only one async endpoint, action defaults to it. If multiple endpoints exist, action is required to avoid querying the wrong task route.'