Maritime-mcp-server defines 12 tools with complete input schemas and descriptions. Naming follows verb_noun convention (vessels, vessel, vessels_near, etc.), though most are noun-only rather than action verbs, acceptable for query/discovery tools. All tools have descriptions (avg 85 chars, within baseline 34 - 392). Input parameters are typed and described. However, output schemas are NOT documented, LLMs cannot see what fields to expect from responses. Error handling is absent; no guidance on retryability, user-fixable errors, or recovery paths. No tool annotations (readOnlyHint/destructiveHint). The set_regions tool is marked WRITE but lacks confirmation/dry-run pattern. Descriptions are functional but terse; they state WHAT but not WHEN or WHY to use each tool. Parameter descriptions are minimal (e.g., 'Latitude of the center point' lacks range/precision guidance). Overall: solid schema foundation, but missing output documentation and error guidance that production tools require.

• Document output schemas for all tools. Specify fields, types, and ranges. E.g., 'Returns array of {mmsi: int, name: string, lat: float (±4 decimals), lon: float (±4 decimals), type: string, status: string, speed_knots: float, distance_nm: float}'. This enables LLMs to extract and chain results.
• Add error handling guidance to tool descriptions. E.g., 'Returns None if vessel not found (not retryable, MMSI does not exist). Try vessels() or vessels_by_type() to discover valid MMSIs.' This guides agent recovery.
• Implement confirmation pattern for set_regions. Add a dry_run parameter or a separate preview_regions tool. Return 'Current: [region1, region2]. Proposed: [region3, region4]. Confirm?' to prevent accidental subscription changes.
• Enhance parameter descriptions with constraints. E.g., 'lat: Latitude in range [-90, 90]. Precision: ±4 decimals (~11m). Required.' and 'radius_nm: Search radius in nautical miles, 1 - 500. Default: 10.' This prevents invalid input.
• Expand tool descriptions to clarify intent and differentiation. E.g., 'Use vessels_near() to find ships around a location (e.g., near a port). Use vessels_by_type() to find all tankers globally. Use vessels_by_status() to find anchored or underway vessels.' This helps LLMs select the right tool.