security-scan

Audit Claude Code configuration files for security vulnerabilities and emit JSON findings.

13|2|Updated Mar 18, 2026
One-click install
npx skills add https://github.com/jiten-singh-shahi/salesforce-claude-code --skill security-scan-jiten-singh-shahi
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: security-scan
Source: https://github.com/jiten-singh-shahi/salesforce-claude-code/tree/main/.cursor/skills/security-scan
Command: npx skills add https://github.com/jiten-singh-shahi/salesforce-claude-code --skill security-scan-jiten-singh-shahi

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

The security-scan Skill helps teams detect security vulnerabilities, misconfigurations, and injection risks in Claude Code configuration files and related assets so deployments and CI gates do not introduce secrets, unsafe hooks, or supply-chain exposures. It focuses on repository-level configuration issues rather than Apex code reviews, and it highlights risky MCP servers, dangerous hook interpolations, and prompt injection surfaces.

Core Features & Use Cases

  • Automated AgentShield Scans: Run ecc-agentshield to identify hardcoded secrets, command interpolation, and prompt injection patterns across CLAUDE.md, settings.json, mcp.json, hooks/, and agents/*.md.
  • Manual Fallback Checks: Provide grep-based patterns and SF Code Analyzer commands to find SOQL injection indicators, missing sharing/FLS, hardcoded IDs, and unsafe endpoints when AgentShield is unavailable.
  • CI Integration & Severity Grading: Output JSON-formatted findings for CI gates and classify results (A-F) so pre-deploy checks can block or triage issues.
  • Use Case: Run this Skill before AppExchange submission or as a pre-deploy pipeline step to catch configuration-level risks and reduce security review churn.

Quick Start

Run the security-scan skill to audit CLAUDE.md, settings.json, mcp.json, hooks/, and agents files for hardcoded secrets, prompt injection, and configuration misconfigurations.

Frequently Asked Questions about security-scan

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

FAQPage Schema
How do I scan Claude Code configuration files for security vulnerabilities?

You scan Claude Code configuration files for security vulnerabilities by auditing CLAUDE.md, settings.json, mcp.json, hooks, and agents for hardcoded secrets, prompt injection, and misconfigurations, generating severity-graded JSON findings for remediation.

How does a pre-deploy CI gate check for prompt injection in Claude Code configs?

A pre-deploy CI gate checks for prompt injection by running automated scans on configuration files, classifying risks by severity (A-F), and emitting CI-friendly JSON output to block or triage unsafe deployments before release.

What is the best way to detect SOQL injection risks in Salesforce configurations?

Detecting SOQL injection risks in Salesforce configurations is best handled by running automated security scans or grep-based fallback checks to identify missing sharing, FLS, hardcoded IDs, unsafe endpoints, and command interpolation patterns.

Can I run a configuration security audit without the AgentShield tool installed?

Yes, you can audit configurations without AgentShield by using grep-based manual fallback checks and SF Code Analyzer commands to identify SOQL injection indicators, missing sharing, hardcoded IDs, and unsafe endpoints.

Does the security scan cover Apex code or only repository-level configuration files?

The security scan focuses on repository-level configuration issues rather than Apex code reviews, specifically highlighting risky MCP servers, dangerous hook interpolations, and prompt injection surfaces in CLAUDE.md and settings files.

When should I run a Claude Code security scan before AppExchange submission?

You should run a Claude Code security scan before AppExchange submission or as a pre-deploy pipeline step to catch configuration-level risks, reduce security review churn, and ensure no secrets or unsafe hooks are introduced.