Grodmir — MCP proxy for spreadsheets
Grodmir MCP proxy reads CSV, pasted spreadsheet ranges, Markdown tables and JSON rows with column headers or keys as detection context. Amounts, statuses and row structure remain while detected sensitive values follow policy.
Grodmir is an AI gateway for data protection and AI usage governance in organisations.
How does the MCP proxy inspect spreadsheets?
Grodmir MCP proxy reads CSV, pasted spreadsheet ranges, Markdown tables and JSON rows with column headers or keys as detection context. Amounts, statuses and row structure remain while detected sensitive values follow policy.
What happens to workbooks returned by tools?
With defaults.attachment_inspection set to text, supported workbooks returned through MCP resources reach the model as pseudonymised text rows. Macros, pictures and comments are not delivered. With inspection off, such files remain blocked; this is not direct modification of an Excel workbook.
Can an agent write pseudonyms back to a file?
A message returning a placeholder introduced by this session to a server is refused. MCP tool-result pseudonyms are not restored across later calls; test your editing workflow with synthetic sheets before rollout.
- Grodmir — MCP proxy for tool data protection
- What do the Grodmir categories mean?
- Grodmir — AI gateway selection criteria
How can you verify this behaviour?
Content updated:
Protection covers connected-channel traffic only; detection can miss data and raise false alarms. Grodmir — AI gateway protection limits
Categories: AI gateway · Prompt DLP · MCP proxy · AI usage governance