Leaks are caught before they leave.
On Enterprise, prompts and MCP tool calls are scanned before they go anywhere, so a credential pasted into a prompt never reaches the model.
DLP on prompts and tool calls (Enterprise)
On Enterprise, prompts and MCP tool calls are scanned before they go anywhere, so a credential pasted into a prompt never reaches the model.
TimeWhoWhereFound
09:31:02Claude Code as mayapromptAWS access keyBlocked
09:44:19Codex as sammcp.callStripe secret keyBlocked
- Checks
- Agent prompts and MCP tool calls
- Patterns
- Built-in credentials, plus your own
- Not covered
- Files and output inside a sandbox are not scanned
- Audit
- Every catch recorded
How it works
What happens, in order.
01Scan
Each prompt and MCP tool call is checked against built-in and your own patterns.
02Block
A credential or other sensitive value stops the call.
03Record
What was found, and where, is in the audit log, never the value itself.
Which plans
DLP comes with Enterprise.
See pricing →Give every agent
a box of its own
Start free with 100 vCPU-hours a month. Upgrade when your agents need more.
A Firecracker microVM for every runNo network by defaultPause and resume with state keptVault keeps secrets out of promptsSSO, roles and audit logBring your own cloudWorks over MCP and the CLI