Replace sensitive patterns with standardized markers in workspace files.
npx skills add https://github.com/0xUrsanomics/utopia-os --skill redact-0xursanomics
Please help me install this Agent Skill. Skill: redact Source: https://github.com/0xUrsanomics/utopia-os/tree/main/skills/redact Command: npx skills add https://github.com/0xUrsanomics/utopia-os --skill redact-0xursanomics
This skill prevents accidental exposure of credentials, PII, or sensitive information by providing a controlled, logged, and reversible redaction process for your memory and output files.
Use the redact skill to remove the sensitive API key from the file located at memory/secrets.md with the reason that it was leaked.
High-intent search queries and answers about installing and using this skill.
To sanitize sensitive data, this skill replaces matched patterns with standardized redaction markers in specific workspace files while logging all modifications to a persistent JSONL audit trail.
Yes, the redaction process enforces a safety-first workflow that requires mandatory user confirmation before any file modification occurs, preventing accidental data loss.
You maintain audit logs for compliance by using a redaction workflow that automatically records every sanitization event in a persistent JSONL audit log, tracking what was scrubbed and why.
No, you cannot redact restricted system files because the skill enforces explicit exclusions of restricted system files to maintain operational safety and prevent critical data loss.
The best way to remove accidentally committed credentials is to run a controlled redaction process that replaces the sensitive values with standardized markers and logs the event for security tracking.
The sanitization process is designed to be reversible by maintaining a persistent JSONL audit log of all modifications, allowing you to track and restore the original sensitive information if needed.
eykdAutomate PII and sensitive data redaction from logs using regex patterns.
MassimilianoPiliAggregate JSONL audit event logs into human-readable Markdown reports.
matteocervelliScan codebases for secrets, vulnerable dependencies, and insecure patterns.
jgwillExtract regex-matching narrative blocks from JSONL message content.
