ContentEngine MCP server has 8 tools with complete input schemas and descriptions, but quality is inconsistent. Tool names follow verb_noun convention (ingest, schedule, fire, cancel, status, list_pending, list_scheduled, sync), which is good. However, descriptions are brief (avg ~80 chars) and lack context about when to use each tool or what happens on failure. Parameters are typed and mostly described, but lack constraints (enums, ranges, format specs). Output schemas are not documented, LLMs cannot predict response structure. Error handling is absent from descriptions. The 'fire' tool (irreversible publish) has no confirmation or dry-run pattern. Overall: functional but below production baseline.

• Document output schemas for all 8 tools. E.g., 'ingest returns {post_id: int, state: string, created_at: ISO8601, source_file: string|null}'. This enables LLMs to chain tools and extract required fields.
• Add enum constraints to 'platform' parameter: platform: {type: 'string', enum: ['linkedin', 'twitter', 'blog']}. Add range to 'priority': priority: {type: 'integer', minimum: 0, maximum: 100}.
• Expand 'fire' description to include: 'Immediately publishes a post, bypassing the queue. This is irreversible, use only when you are certain the post is ready. Returns error if post is already published or in invalid state.'
• Add pagination to list_pending and list_scheduled: limit (1-100, default 20), offset (default 0), and return {items: [...], total: int, has_more: bool}. Cap results at 50 items per call.
• Add error recovery guidance to descriptions. E.g., 'schedule' should note: 'If scheduled_at is in the past, returns error. Use current time or future time. If post is already scheduled, call cancel first.'

Inferred effective spec: <=2025-11-25.

• Relies on Logging (deprecated) - log to stderr or use OpenTelemetry