ERPipe has 29 tools with mostly complete schemas and descriptions, but quality is inconsistent. Naming follows verb_noun convention well (search_records, read_record, execute_approved_write). Descriptions exist for all tools but many are terse (10-50 chars) and lack context on WHEN to use them or what happens. Parameter descriptions are sparse, most params lack explanations of format, constraints, or dependencies. Output schemas are not documented. Error handling guidance is absent. The approval token workflow (preview_write → validate_write → execute_approved_write) is well-composed, but many tools lack idempotency hints and recovery guidance.

• Add parameter descriptions to all tools. For each param, explain format (e.g., 'Odoo domain syntax: [[field, operator, value], ...]'), constraints (e.g., 'limit: 1-50, default 10'), and when it's required vs optional. Use the pattern: 'The [field] to [action]. Format: [example]. Range: [min-max].'
• Document output schemas for all tools. Add a 'Returns' section to each description listing the structure of the response object, field names, types, and which fields are safe to pass to downstream tools. E.g., search_records should document that it returns {records: [{id, name, ...}], total_count: number}.
• Expand terse descriptions to 50-150 characters. Include WHAT the tool does, WHEN to use it (vs similar tools), and any prerequisites. E.g., 'List Odoo models matching a query. Call this first to discover available models before calling get_model_fields. Supports optional name filtering and result limits.'
• Add error handling guidance to state-modifying tools. Describe failure modes (e.g., 'Token expired', 'Record locked', 'Permission denied'), classify as retryable/user-fixable/fatal, and suggest recovery steps. E.g., 'If token is invalid, call validate_write again to get a fresh token.'
• Clarify idempotency and side effects. For execute_method, chatter_post, and execute_approved_write, state whether repeated calls with the same input produce the same result or have cumulative effects. This guides agent retry logic and prevents duplicate actions.