- Home
- Skills
- Code Quality & Review
- AI Generated Code Verification Engineer Skill Pro
Works with the AI Agents you already use
AI Generated Code Verification Engineer Skill Pro
Independent verification engineer that audits code via requirement matrices, test execution, and security checks.
$19.90
AI Generated Code Verification Engineer Skill Pro
Example session with this skill installed
Verify the latest changes in the auth-service directory. Check for SQL injection risks and ensure the new JWT rotation logic has passing integration tests. Work in read-only mode.
- Read your context and instructions
- Compiled the ai generated code
I have completed the verification of the auth-service changes focusing on the JWT rotation logic and SQL injection risks.
AI-Generated Code Verification Report
Executive verdict
- Verdict: VERIFIED WITH RISKS
- Scope:
services/auth-service/** - Overall confidence: Medium
- Merge recommendation: Approve with conditions (address Medium finding below)
Requirements traceability
| Requirement | Evidence | Status |
|---|---|---|
| Implement JWT rotation | integration/jwt_test.go | PASS |
| Prevent SQL Injection | Static analysis of db/user_repo.go | PASS |
| Secure secret handling | Code review of environment loading | PARTIAL |
Checks executed
go test ./services/auth-service/integration/...: PASS (Exit 0)gosec ./services/auth-service/...: PASS (Exit 0)golangci-lint run ./services/auth-service/...: FAIL (Exit 1 - 2 style issues)
Findings
- Severity: Medium
- Title: Potential Secret Exposure in Logs
- File:
services/auth-service/auth.go:88 - Evidence:
log.Printf("failed to rotate for token: %s", token) - Impact: Refresh tokens may be leaked to centralized logging systems.
- Remediation: Use a hashed identifier or mask the token string before logging.
Gaps and limitations
- No load testing performed on the rotation endpoint.
- Hardware Security Module (HSM) integration was mocked in tests.
Next steps
- Mask the token value in the failure log at
auth.go:88. - Run the full regression suite for the
identity-providermodule. - Approve the PR once the logging fix is verified.
Connects securely to your tools. The creator never sees your data.
About this skill
AI-generated code often suffers from "hallucinated correctness" where scripts look plausible but fail in production or harbor hidden security flaws. This skill provides an independent, evidence-based verification layer to ensure AI or human contributions actually do what they claim. It replaces blind trust with a formal audit process, requiring concrete proof for every pass status.
What it does
- Requirement Traceability maps each implementation change directly to stated acceptance criteria to ensure zero feature drift.
- Evidence-Based Auditing generates a requirement-to-evidence matrix that separates observed facts from unverified claims.
- Automated Check Discovery inspects your
package.json,README, or CI configs to run the project's actual test suites and linters. - Security-Focused Review identifies high-risk patterns like SQL injection, unsafe secret handling, and broken access controls.
- Verification Reporting produces a structured verdict (Verified, Rejected, or Verified with Risks) backed by command output and exit codes.
How it works
- Scope Baseline: The agent inspects the diff, repository status, and manifests to understand the change set.
- Matrix Construction: It converts requirements into a matrix of verifiable claims and negative test cases.
- Check Execution: It runs unit, integration, and static analysis checks using the project's own toolchain.
- Report Generation: It delivers a final report including the executive verdict, executed commands, and prioritized findings.
Frameworks & tools
The skill is language-agnostic and adapts to the tools found in your repository, including unit testing frameworks, linters, static analyzers, and dependency scanners.
Why this beats prompting it yourself
Generic prompts often lead to the AI agreeing with itself or assuming code works because it "looks right." This skill enforces a skeptical operating mode that demands execution evidence, exit codes, and negative testing before issuing a verdict.
Use cases
- AI Agent Supervision to audit the output of coding agents before they reach your main branch.
- CI/CD Gatekeeping to determine if current test coverage is sufficient for a defensible release.
- Security Triage to quickly scan new PRs for injection risks or sensitive data leaks.
- Refactoring Validation to ensure state transitions and edge cases remain intact after logic changes.
Known limitations
Does not perform live system exploitation. Requires explicit authorization for network access or modifying production code. Limited by the tools and test suites available in the local environment.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Code Quality & Review.
Cannot perform live exploits. Requires user permission for network/file writes. Verdict accuracy depends on the quality of local test suites and available static analysis tools.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 16 days ago