Secret Redaction

Redact secret-bearing input before it enters the model context. Use this before opening .env*, credential files, token files, or command output that may print secrets.

Inputs

• One or more files that may contain secrets.
• A raw value, command output, or pasted text that may contain secrets.
• The verification goal: present keys, empty values, missing values, or whether two values match.
• The format check goal: whether a value matches a common credential shape.

Workflow