Tool Output Privacy Firewall
by Enrique brachoUpdated Sep 1, 2026
Redact secrets and PII and minimize MCP tool outputs before they reach an AI model.
mcp
ai agents
privacy
+7
|How to pay
Subscribe
Monthly billing
$9/month
Predictable monthly cost with included usage. Best for steady, high-volume traffic.
- Unlimited tools within plan limits
- One API key, billed once a month
- Cancel any time
Tool Output Privacy Firewall adds a deterministic privacy layer between MCP tool results and an AI model.
It can:
- Redact API keys, authentication tokens, passwords, private keys and database credentials.
- Detect and redact emails, phone numbers, Social Security numbers and payment cards.
- Keep only explicitly approved JSON fields.
- Remove denied fields and unnecessary data.
- Inspect outputs and report privacy risks without exposing the original sensitive values.
The firewall does not use an external AI model, make outbound network calls or persist submitted tool outputs.
Available tools:
- sanitize_tool_output
- inspect_tool_output
- minimize_tool_output
This is a defensive privacy utility, not a compliance certification or a guarantee that every possible sensitive value will be detected.