What happens if the scanner finds no secrets?

If the initial scan returns zero candidates, the orchestrator will skip the analysis phase and provide a summary report confirming that no secrets were detected.

What types of credentials can this skill detect?

The Ghost Secrets Scanner detects a wide range of sensitive data including leaked API keys, tokens, passwords, and exposed environment variables in your codebase.

Does it provide remediation steps for found secrets?

Yes, every finding includes a severity level and specific remediation guidance to help you properly rotate or secure the exposed credential.

How does the orchestration process work?

The skill acts as a top-level orchestrator that spawns subagents to handle specific tasks: initializing the environment, scanning files, analyzing results, and summarizing the findings.

Is the scanning performed locally?

The skill initializes and runs the Poltergeist binary on your local machine, allowing for secure scanning of your repository directly through the Claude Code interface.

Ghost Security Secrets Scanner

Name: Ghost Security Secrets Scanner
Author: ghostsecurity

byghostsecurity

•

357

•

セキュリティとテスト

Scans codebases to identify and remediate hardcoded secrets, API keys, and sensitive credentials using automated security agents.

Ghost Security Secrets Scanner is a specialized AppSec skill for Claude Code designed to automate the detection of leaked credentials within your repository. By orchestrating a chain of specialized subagents, the skill initializes the Poltergeist scanner, identifies candidate secrets, analyzes them for confirmation, and generates a comprehensive report. It is an essential tool for developers looking to prevent sensitive data exposure, providing clear severity levels and actionable remediation guidance for found vulnerabilities.

主な機能

01Specific remediation guidance for each identified secret

02Detailed findings reports with severity assessments

03357 GitHub stars

04Automated detection of API keys, tokens, and passwords

05Multi-agent orchestration for high-accuracy verification

06Integrated support for Poltergeist security binary

ユースケース

01Automating security reviews within the Claude Code terminal environment

02Auditing a repository for hardcoded credentials before a public release

03Scanning configuration files and scripts for accidentally committed secrets

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add ghostsecurity/skills scan-secrets

For use in Claude.ai and ChatGPT

Download Skill