kanka-mcp has clear verb-based naming (show_, list_, get_, create_, update_, delete_) and complete input schemas with types. However, descriptions are inconsistent in depth. Tool descriptions range from minimal (e.g., 'List all characters in a campaign') to verbose (create_character with 15-line field documentation). Parameter descriptions are present but generic, many lack format constraints, ranges, or examples of valid values. No output schema documentation. Error handling is absent: tools call raise_for_status() but provide no recovery guidance. No pagination support despite list_* tools. Security: API key is injected via environment variable (correct), but no permission checks or audit logging. Composition is reasonable, tools are single-purpose, but list_characters and list_locations lack limit/offset parameters, risking context window overflow.

• Add output schema documentation for all tools. Example: 'Returns: {campaigns: [{id: int, name: string, ...}], total: int}'. This enables LLMs to chain tools correctly.
• Add pagination to list_characters and list_locations: accept limit (1-100, default 20) and offset parameters; return total_count and next_offset in response.
• Wrap HTTP errors in actionable recovery guidance. E.g., catch requests.exceptions.HTTPError and return '{"error": "Campaign not found. Verify campaign_id with show_campaigns() first."}'
• Add a confirm_delete_character tool or require a confirmation_token parameter in delete_character to prevent accidental deletion.
• Document parameter constraints in descriptions: 'age (string, e.g., "25" or "young adult")', 'tags (comma-separated string or array of tag IDs)', 'entry (HTML string; plain text will be wrapped in <p> tags)'.