Will it interrupt my flow for simple tasks?

No, the skill is specifically designed not to suggest verification for trivial fixes, simple renames, CSS styling, or exploratory coding.

Can I decline the verification suggestion?

Yes. If you decline a suggestion, the skill will acknowledge your choice and proceed directly without repeating the suggestion for similar questions in that session.

When will the skill prompt me for verification?

It prompts during tasks involving significant complexity, such as async promise chains, security-sensitive operations, financial logic, or ambiguous bug investigations.

SE-CoVe stands for Software Engineering Chain-of-Verification, a methodology designed to systematically validate claims and logic within software development to reduce subtle bugs.

Does this skill run verification automatically?

No, the skill follows a minimal and non-eager philosophy. It only suggests using the /verify command when it detects high-risk scenarios, leaving the final decision to the developer.

Software Engineering Auto-Verify

Name: Software Engineering Auto-Verify
Author: vertti

byvertti

•

보안 및 테스팅

Prompts systematic verification for complex software engineering tasks to ensure code reliability, security, and logical correctness.

This skill implements the Software Engineering Chain-of-Verification (SE-CoVe) methodology by identifying high-risk or complex coding scenarios and suggesting the use of the /verify command. It serves as a non-intrusive safety net, helping developers catch subtle bugs in asynchronous logic, state management, and security-sensitive operations by prompting for a deeper analytical review only when a task meets specific complexity thresholds. By focusing on verifiable claims and high-value scenarios like race conditions or architectural tradeoffs, it ensures that critical code is thoroughly vetted without slowing down routine development tasks.

주요 기능

01Context-aware verification prompts for high-complexity code

02Security-focused auditing for authentication and data handling

03Specialized detection for async patterns and race conditions

04Systematic evaluation of complex architectural tradeoffs

05Non-intrusive, user-first design that never auto-invokes

0619 GitHub stars

사용 사례

01Investigating intermittent production bugs where the root cause is uncertain

02Hardening security-sensitive operations like token handling or encryption

03Implementing complex React hooks or state management logic with potential edge cases

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add vertti/se-cove-claude-plugin auto-verify

For use in Claude.ai and ChatGPT

Download Skill