Server has 14 tools with consistent naming (verb_noun pattern) and explicit schemas. Most tools have descriptions (10-150 chars), but many are terse and lack actionable context. Parameter descriptions are present but minimal. Output schemas are not documented. Error handling exists but lacks recovery guidance. No tool annotations (readOnlyHint/destructiveHint). Composition is reasonable, tools are single-purpose, but chaining metadata (IDs needed for follow-up calls) is inconsistent.

• Add output schema documentation to every tool. Document the structure of returned data (e.g., glif object has {id, name, description, inputs: [{name, type}]}, pagination has {total, cursor}). Use JSDoc or inline comments in handler functions.
• Expand tool descriptions to 50 - 150 characters. Include WHAT the tool does, WHEN to use it vs similar tools, and any prerequisites. E.g., 'Get all glifs you own. Use this to list your personal workflows; use search_glifs to find public or shared glifs.'
• Add tool annotations to the definition object. Mark read-only tools with readOnlyHint: true, destructive tools with destructiveHint: true, and idempotent tools with idempotentHint: true. This guides agent caution and retry logic.
• Enhance error messages with recovery guidance. Instead of 'Error: <message>', return structured errors like {error: 'not_found', message: 'Glif not found', suggestion: 'Try search_glifs() to find available glifs'}. Categorize as retryable, user-fixable, or fatal.
• Expand parameter descriptions with format and constraint hints. E.g., 'prefix: Optional prefix for tool names (alphanumeric and underscores only, max 20 chars). Example: "my_bot_"'. Use enums for constrained values and document what each enum value means.