Spring Boot MCP server with Spring AI integration for marketing campaign assistance. Provides MCP server and client functionality with Azure OpenAI integration for AI-powered marketing campaign management.
This MCP server has critical definition quality issues across all 6 tools. All tools share an identical, severely under-specified input schema with only a single 'request' parameter of type string. No output schemas are documented anywhere. Descriptions are generic and lack the specificity needed for LLM tool selection. Tool naming violates verb-noun patterns in 4 of 6 cases (audience_segmentation, channel_strategy, email_template_generation, campaign_performance_analysis, budget_allocation_optimization all use noun-adjective patterns instead of verb_noun). The single string parameter design eliminates all input validation, type safety, and parameter-level descriptions. No evidence of error handling guidance, permission gates, or security controls. Based on rubric calibration, this server lands in the F-D range: definitions are barely functional.
Define and analyze target audience segments
Optimize budget allocation across marketing channels
Create comprehensive marketing campaigns
Analyze campaign performance metrics and KPIs
Develop multi-channel marketing strategies
Generate marketing email templates and content
All 6 tools use identical single-string input schema with 'request' parameter only. No structured input validation, no parameter-level type safety, no per-parameter descriptions. Violates constrained-input and tool-description patterns.
No output schemas documented for any tool. LLMs cannot plan downstream calls or know what fields to expect. Violates documented-output and response-shaper patterns.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | F | 34 | 2026-07-28+ | v2 |
| 2026-03-09 | F | 28 | - | v1 |
Tool names violate verb_noun convention. 'audience_segmentation', 'channel_strategy', 'email_template_generation', 'campaign_performance_analysis', 'budget_allocation_optimization' are noun-based. Should be: 'create_audience_segment', 'develop_channel_strategy', 'generate_email_template', 'analyze_campaign_performance', 'optimize_budget_allocation'. LLMs rely on action verbs in names to infer tool intent.
Tool descriptions are generic and under 50 characters. E.g., 'Create comprehensive marketing campaigns', 'Define and analyze target audience segments'. No explanation of WHEN to use each tool vs. alternatives, WHAT it returns, or prerequisites. LLMs cannot differentiate between similar tools.
No error handling guidance or recovery hints. If a tool fails or returns invalid input, LLMs have no direction on what to do next. Violates recovery-guide pattern.
No permission gates or audit logging visible in tool definitions. WRITE-risk tools (5 of 6) should declare required permissions (e.g. 'write:campaigns', 'write:budget') and log who called what and when. Violates permission-gate and audit-trail patterns.