Loading ToolBench
Model Context Protocol (MCP) server for Superlines AI visibility analytics. Connect Claude Desktop, Cursor, or any MCP-compatible client to your Superlines data.
This server has severe definition quality issues across all dimensions. The source code shows only tool name constants and descriptions in a simple object literal (src/index.ts AVAILABLE_TOOLS), with NO visible input schemas, parameter descriptions, output schemas, or error handling patterns. The mcp-proxy.js file is a basic STDIO proxy that forwards JSON-RPC to a remote server but provides no tool definitions locally. Critically, the actual tool implementations are not in the provided source, only names and one-line descriptions. All 22 tools lack documented input parameters, type definitions, output schemas, and implementation details. This violates fundamental MCP tool definition requirements.
Competitor mentions with sentiment
Deep-dive analysis of competitor URLs
Core metrics: visibility, citations, share of voice
Sentiment analysis of AI responses
Topics with improvement potential
High-level analytics overview
Best performing prompts
Get detailed brand configuration and competitors
NO INPUT SCHEMAS VISIBLE. All 22 tools lack documented parameter definitions, types, and constraints. The source code shows only tool names and one-line descriptions in AVAILABLE_TOOLS constant. No JSON Schema definitions present for any tool.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | F | 32 | <=2025-11-25 | v2 |
| 2026-03-09 | F | 22 | - | v1 |
Domain and URL citation analysis
Find competitive opportunities
Comprehensive competitor overview
LLM source query analysis
Compare current vs previous period
Query analysis with search volumes
Top cited URLs per query
Weekly trend analysis
List all brands available to the user
Optimize Schema.org markup
Content quality analysis
Technical SEO analysis
Full LLM-friendliness audit
Fetch and parse webpage content
DESCRIPTIONS ARE GENERIC AND TOO SHORT. Most tool descriptions are 30 - 40 characters ('Weekly trend analysis', 'Query analysis with search volumes'). These lack context about when to use each tool, what parameters are required, what output to expect, and how to chain tools. LLMs cannot disambiguate between similar tools.
NO OUTPUT SCHEMAS DOCUMENTED. The source shows no declared return types, field schemas, or result structures for any tool. Agents cannot plan downstream calls or parse results reliably.
TOOL DEFINITIONS NOT IN SOURCE. The provided src/index.ts contains only a constant listing tool names and descriptions; the actual MCP tool registration (with input schemas, parameter types, and output documentation) is absent. This violates the requirement that tool definitions be visible in source code. The mcp-proxy.js forwards requests to a remote server, meaning tool definitions are hosted remotely and not auditable from this repo.
NO PARAMETER VALIDATION OR DOCUMENTATION. No evidence of parameter constraints (enums, min/max, regex patterns, required vs optional). LLMs will pass arbitrary values, causing failures.
NO ERROR HANDLING PATTERNS. No evidence of error classification, recovery guidance, actionable error messages, or dry-run confirmation for destructive operations (e.g., schema_optimizer is marked WRITE but no undo/confirmation mechanism shown).
SIMILAR TOOL NAMES CREATE AMBIGUITY. Multiple tools analyze different aspects without clear naming distinctions: 'get_query_data' vs 'get_fanout_query_insights' vs 'get_best_performing_prompt', LLMs will struggle to select the right one.
PAGINATION NOT DOCUMENTED. Tools like 'list_brands' and 'analyze_brand_mentions' likely return multiple results but no pagination parameters (limit, offset, cursor) or total count are documented.