MCP Server for WordPress management via ChatGPT using Server-Sent Events (SSE) transport
Server has 4 tools with basic schemas and descriptions, but significant gaps prevent production use. All tools have input schemas with type definitions and descriptions present, meeting minimum requirements. However, descriptions are in Russian (non-standard for LLM optimization), lack actionable context, and omit critical guidance on when to use each tool. No output schemas documented. Error handling returns generic success/failure flags without recovery guidance. Missing tool annotations (destructiveHint critical for delete_post). Parameter descriptions are minimal (avg ~30 chars vs baseline 72). No pagination limits documented for get_posts despite returning arrays. Credentials hardcoded in source (critical security issue). Overall pattern adherence ~25%.
Создать новый пост в WordPress
Удалить пост
Получить список постов
Обновить существующий пост
Credentials hardcoded in source code (WORDPRESS_USERNAME, WORDPRESS_PASSWORD visible in mcp_sse_server.py). Violates secret-injection pattern, credentials must use environment variables or vault.
delete_post lacks destructiveHint annotation and confirmation mechanism. No dry-run or user confirmation before permanent deletion. Agents can delete posts without safeguards.
All tool descriptions in Russian ('Создать новый пост в WordPress', 'Удалить пост'). LLMs are optimized for English descriptions. Non-English descriptions reduce tool selection accuracy and token efficiency.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 50 | <=2025-11-25 | v2 |
No output schemas documented. Callers cannot predict response structure (success flag, post_id, url, message fields). LLMs must infer structure from examples, increasing hallucination risk.
get_posts returns formatted_posts array without pagination limit enforcement in response. Code caps per_page at 100, but no description warns agents of this limit or advises pagination strategy.