Question 1

Is it safe to paste passwords or API keys into CleanMyPrompt?

Accepted Answer

Yes. All processing is client-side by default; the app does not send pasted content to our servers. We also detect and redact common API key and secret patterns (AWS, OpenAI, GitHub). For extra safety, enable 'Audit' mode to preview detected secrets before accepting changes.

Question 2

How can I reduce Claude or Gemini API costs?

Accepted Answer

Select the correct model in the dropdown, then use Token Compression. For Claude, prefer conservative compressions and validate outputs. See the Token Compression guide in the footer for best-practice links.

Question 3

Can I remove Social Security Numbers (SSNs) or other PII?

Accepted Answer

Yes. The PII Scrubber detects common national ID formats and redacts them. After running, review the 'Audit' results; we provide a visual diff so you can confirm removals. For regulatory workflows, always verify before sharing.

Question 4

Is CleanMyPrompt HIPAA / healthcare safe?

Accepted Answer

CleanMyPrompt is designed for privacy-first use: processing happens in your browser. HIPAA compliance depends on your organization's processes — contact your privacy officer. See our Privacy page for implementation notes.

Question 5

How does Token Squeeze work?

Accepted Answer

Token Squeeze removes boilerplate, markdown links, and stop-words. In aggressive mode it removes low-density sentences. Use 'Preview' to inspect the output and the visual diff to understand what's removed.

Question 6

Do you store my text or usage?

Accepted Answer

No text is uploaded unless you opt into anonymous analytics — which only records non-content metrics. Session history (if allowed) is saved locally in your browser. See the Privacy page for details.

Question 7

Is CleanMyPrompt EU AI Act compliant?

Accepted Answer

Yes. CleanMyPrompt uses deterministic regex and NLP rules — not AI models — so it does not fall under high-risk classification. We provide a full transparency page with Article 52, Article 5, and Article 13 compliance documentation. Visit our Transparency page for details.

Question 8

Can AI agents use CleanMyPrompt as a tool (MCP)?

Accepted Answer

Yes. CleanMyPrompt provides an MCP-compatible tool definition at /mcp.json and an OpenAPI spec at /openapi.json. AI agents can call clean_text and analyze_text tools via the REST API — no API key required. See our API docs for integration examples.

Question 9

Can I blur PII in screenshots or images?

Accepted Answer

Yes. Upload an image and use the built-in PII Blur tool to draw rectangles over sensitive areas. The blur uses pixelation (not overlay) so the data is permanently destroyed. After blurring, the image is sent through OCR to extract text — all in your browser.

Question 10

Does the REST API store my text?

Accepted Answer

No. The API processes text server-side then discards it immediately after returning the response. No logs, no storage, no training. The server retains zero data. For maximum privacy, use the browser-based tool which never sends data to any server.

Feature	CleanMyPrompt	Manual Cleaning	Microsoft Presidio	Commercial Tools
Price	Free forever	Free (your time)	Free (self-host)	$50-500/mo
Setup Required	None — open & paste	None	Python + Docker	API key + SDK
Data Privacy	100% in-browser	Depends	Self-hosted only	Cloud upload
Token Compression	Up to 40%	Manual effort	No	Varies
AI Agent / MCP Ready	Yes — REST API + MCP	No	API only	API only
Multi-modal (Images)	OCR + PII Blur	No	Text only	Some
Global Compliance Ready	GDPR · CCPA · HIPAA · AI Act	N/A	Self-assess	Varies
No Account Needed	Zero sign-up	✓	Server setup	Required
Browser Extension	Chrome — free	No	No	Rarely

The Definitive Shadow AI Firewall.

Clean in three steps

Paste Your Prompt

Scan & Detect

Copy Clean Prompt

Who needs CleanMyPrompt?

Developers

Legal & HR

Healthcare

Prompt Engineers

Why CleanMyPrompt?

Clean prompts directly inside ChatGPT, Claude & Gemini