Dropbox integration for file management, search, and content operations
Dropbox MCP server has well-structured tool definitions with clear naming conventions and comprehensive parameter descriptions. All four tools follow verb_noun naming patterns (list_files, read_file, search_files, write_file). Parameter schemas include type definitions and actionable descriptions. However, output schemas are not formally documented in the source code, they are inferred from implementation. Error handling exists but lacks explicit recovery guidance for common failure modes. The server demonstrates good composition patterns with tools chained appropriately (search_files → read_file), though some advanced patterns like idempotency hints and confirmation steps for destructive operations are absent.
List files and folders in a Dropbox directory. Use this to browse and explore the Dropbox folder structure. Returns name, path, type, size, and modification date for each entry.
Read the contents of a text file from Dropbox. Supports markdown, text, code, JSON, YAML, CSV, and other text-based files. Binary files (images, PDFs, etc.) will return metadata only.
Search for files in Dropbox by name or content. Use this to find files when you don't know the exact path. Searches file names and, where possible, file contents.
Create or update a text file in Dropbox. Writes UTF-8 text content to the specified path. Parent folders are created automatically.
Output schemas not formally documented. Tool descriptions explain what is returned (e.g., 'Returns name, path, type, size, and modification date') but the response format is inferred from code only. LLMs cannot plan multi-step chains reliably without explicit output schema declarations.
write_file lacks destructive operation safeguards. Tool modifies user data (creates/overwrites files) but has no dry-run mode, confirmation step, or audit trail integration. Description does not explicitly warn that this is irreversible.
Error handling does not guide LLM recovery. Source code shows try/except blocks that catch dropbox.exceptions but the error messages returned to the LLM are not visible. Cannot verify that errors include actionable recovery hints (e.g., 'File not found. Try search_files() first.').
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | C | 66 | 2026-07-28+ | v2 |
| 2026-03-09 | F | 0 | - | v1 |
read_file description does not specify maximum file size or behavior for binary files. Docstring mentions 'Binary files (images, PDFs, etc.) will return metadata only' but does not clarify what 'metadata' includes or the size limit for text files.
search_files parameter 'file_extensions' accepts a comma-separated string but lacks constraint validation visible in schema. Could benefit from enum alternatives or explicit format documentation to reduce invalid LLM inputs.