security-fix

Run parallel security reviews and fix critical findings with verification.

Updated Mar 3, 2026

Search Tags:#automation #security #vulnerability #ci-cd #code-review #patching

Ready to run

npx skills add https://github.com/sukeun1997/claude-config --skill security-fix-sukeun1997

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: security-fix
Source: https://github.com/sukeun1997/claude-config/tree/main/skills/security-fix
Command: npx skills add https://github.com/sukeun1997/claude-config --skill security-fix-sukeun1997

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

End-to-end security review and fix automation that runs code-reviewer and security-reviewer in parallel, filters CRITICAL and HIGH findings, and orchestrates automated fixes, verification, and commit.

Core Features & Use Cases

Parallel reviews: code-reviewer and security-reviewer run concurrently, surfacing high-severity issues for rapid remediation.
Phase-driven workflow: Review → Evaluate → Fix → Verify → Commit, with evidence-based validation and user confirmation before executing fixes.
Use Case: When preparing a project for a security audit or regulatory check, run this skill to identify vulnerabilities, apply minimal patches, and verify a clean build and test before committing.

Quick Start

Invoke the security-fix skill to start a full security review and fix pipeline across the specified path.

Frequently Asked Questions about security-fix

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

FAQPage Schema

How do I automate vulnerability patching and security review in my CI-CD pipeline?▼

Automated vulnerability patching runs parallel code-reviewer and security-reviewer processes to filter CRITICAL and HIGH findings, apply minimal fixes, and verify builds. This pipeline scopes entire repositories or targeted directories to remediate issues before committing changes.

What is the best way to fix high severity security findings without introducing breaking changes?▼

Fixing high severity findings requires applying strict minimum-diff changes to the affected code. The workflow enforces evidence-based validation and mandatory user confirmation before executing automated patches, ensuring fixes do not introduce regression or break existing functionality.

Can I run an end-to-end security fix workflow on a specific directory instead of the whole repository?▼

End-to-end security fix workflows support scoping targeted directories as well as entire repositories. You can invoke the review and fix pipeline across a specified path to isolate critical vulnerabilities and apply patches without scanning unrelated code.

How does automated security verification work after applying code patches?▼

Automated security verification follows a structured phase-driven workflow that validates patched code using evidence-based testing. After applying fixes, the pipeline executes verification checks to confirm a clean build and test suite before proceeding to the final commit phase.

Do I need to manually confirm security fixes before the pipeline commits changes to my repository?▼

Manual confirmation is mandatory before the pipeline proceeds to the fix execution phase. The workflow pauses for user approval after evaluating CRITICAL and HIGH findings, ensuring you maintain control over automated patching and subsequent commits.

What are the limitations of using automated security review for regulatory audit preparation?▼

Automated security review for regulatory audits focuses strictly on CRITICAL and HIGH severity findings. It filters out lower severity issues and requires mandatory user confirmation at Phase 3, meaning it does not provide a comprehensive low-severity compliance checklist without manual supplemental review.