Everything ScrubMyText can do, in one place.
Choose a tool below. Browser tools run locally on your device; the same capabilities are also exposed through REST and MCP for automated workflows.
Scrub text
Deterministically normalize hidden/control characters and copy/paste formatting artifacts.
Inspect text
Find exact Unicode and invisible-character anomalies without modifying the text.
Scan untrusted text
Look for structural risk signals such as invisible controls, mixed scripts, encoding, and instruction-override patterns. This is not a maliciousness verdict.
Redact sensitive text
Replace common secret and personal-identifier patterns while preserving the rest of the text.
Compare text
Verify exactly whether two versions differ and whether differences are only standard scrub artifacts.
Detect secrets
Find likely API keys, private keys, JWTs, cloud credentials, and related secret patterns without echoing the values in the report.
Validate AI output
Parse JSON and validate the common JSON Schema rules most agent tool calls rely on.
Sanitize agent context
Normalize untrusted text and optionally redact sensitive patterns before it is passed to a model.
Prepare for agent
Run the full ScrubMyText preprocessing pipeline in one local browser action.
Using these from an AI agent?
The REST API and remote MCP server expose the same tool names with structured JSON results.
Developer docs →Want one preprocessing call?
prepare_for_agent combines the core checks and returns agent-ready text plus a deterministic report.