Four tools with basic schemas and descriptions, but significant gaps in parameter documentation and output schema clarity. Tool names follow verb_noun convention (health_check, extract_text, ocr_pdf, extract_text_with_ocr). Descriptions exist but are brief (34-80 chars, below the 194-char baseline). Input schemas present with type info, but parameter descriptions are minimal. No output schemas documented. Error handling is implicit in code but not surfaced in tool definitions. The extract_text_with_ocr tool combines two concerns (extraction + OCR), violating single-responsibility principle.
Extract text from a PDF file with structure detection (headings, paragraphs, tables).
Extract text from a PDF file, applying OCR if needed.
Verify PDF MCP server and OCRmyPDF installation. Returns: "✅ PDF MCP server is healthy" or "❌ [error message]"
Apply OCR to a PDF file using OCRmyPDF to extract text from scanned documents.
No output schemas documented. Tool definitions lack return type specifications, forcing LLMs to guess what fields are available for downstream tool calls or response parsing.
extract_text_with_ocr combines two responsibilities (text extraction + OCR application). Should be split into separate tools so agents can compose them independently.
Parameter descriptions are minimal or missing context. 'pdf_path' lacks guidance on absolute vs relative paths, file size limits, or supported PDF versions. 'force_ocr' lacks explanation of when to use it.
Tool descriptions are too brief (34 - 80 chars vs 194-char baseline). health_check description is 34 chars; extract_text is 80 chars. Insufficient context for LLM tool selection.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 48 | <=2025-11-25 | v2 |
No error recovery guidance in tool definitions. Code handles timeouts, missing OCRmyPDF, and invalid page ranges, but these error paths are not documented in tool descriptions to guide LLM recovery.