secrets-tools

Scan repository publish surfaces for secrets and output redacted JSON findings.

5|Updated Nov 26, 2025
One-click install
npx skills add https://github.com/EffortlessMetrics/demo-swarm --skill secrets-tools
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: secrets-tools
Source: https://github.com/EffortlessMetrics/demo-swarm/tree/main/.claude/skills/secrets-tools
Command: npx skills add https://github.com/EffortlessMetrics/demo-swarm --skill secrets-tools

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill addresses the critical security risk of accidental exposure of sensitive information (API keys, private keys) in code or artifacts. It automates the detection and redaction of secrets, preventing costly data leaks and ensuring compliance.

Core Features & Use Cases

  • Secure Scanning: Scans files for common secret patterns (GitHub tokens, AWS keys, private keys, JWTs) and reports locations only, strictly never exposing the secret content itself.
  • In-Place Redaction: Safely replaces identified secret types with [REDACTED:type] placeholders directly within files.
  • Publish Gate Integration: Designed to determine safe_to_publish status, acting as a critical safeguard against accidental secret exposure before deployment or publication.
  • Use Case: Before publishing any artifacts or code, use this skill to scan the relevant files for secrets. If any are found, automatically redact them in-place and report the findings to a JSON file, ensuring no sensitive data leaves the secure environment.

Quick Start

Use the secrets-tools skill to scan the '.runs/my-feature/deploy' directory for secrets and output findings to 'secrets_scan.json'.

Frequently Asked Questions about secrets-tools

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I prevent secrets like API keys from being accidentally committed to my repository?

Secrets scanning automatically detects common secret patterns (GitHub tokens, AWS keys, private keys, JWTs) in your repository's publish surface before deployment, reporting locations only—never exposing the actual secret values—to stop leaks at the gate.

Can I automatically redact secrets found in files before publishing?

Yes. In-place redaction replaces identified secrets with `[REDACTED:type]` placeholders directly in files, letting you sanitize sensitive data safely without manual intervention before artifacts or code leave your secure environment.

What's the best way to integrate secrets detection into a publish pipeline?

Use secrets scanning as a publish gate to determine `safe_to_publish` status. Scan staged files and the repository's publish surface, output findings to JSON, and block deployment if secrets are detected, ensuring no sensitive information reaches production.

Does secrets scanning work with DevSecOps compliance workflows?

Yes. Secrets scanning meets functional requirements for security hygiene, compliance gates, and redaction workflows by operating on the repository root's publish surface through a shim, supporting both 'secrets scan' and 'secrets redact' commands with JSON findings output.

What types of secrets can be detected and redacted?

The tool scans for common secret patterns including GitHub tokens, AWS keys, private keys, and JWTs, reporting their file locations and types while strictly protecting the actual secret content through redaction markers.