Scan before opening a file
Put a CSV or text file inside an allowed folder. Ask Claude Code:Scan customers.csv with DataFog before opening it. Tell me what types of sensitive data you find, without reading the file into the conversation.Use an absolute path if the file is outside the agent’s current directory. The scan returns counts by entity type and finding locations. CSV/TSV findings include record and column numbers; text findings include character offsets. You can try the fictional customer CSV from the MCP repository. Save it inside an allowed folder first.
Create a protected copy
After scanning, ask:Create a redacted copy of customers.csv with DataFog. Use the default checks, leave the original unchanged, and tell me the copy’s path. Do not open either file.The original stays intact. By default, copies are saved beside it; existing copies are never overwritten. Configure a different destination with
datafog-mcp policy --edit.
Detection coverage
Default checks cover email, SSN, credit cards, labeled US routing numbers and NPIs, and supported credentials. An explicitentity_types selection replaces
the defaults. Review coverage and file limits before
choosing which types to include.
For custom output folders, see configuration.