Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    AI Generated Code Verification Engineer Skill Pro

    1

    Independent verification engineer that audits code via requirement matrices, test execution, and security checks.

    $19.90

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    AI Generated Code Verification Engineer Skill Pro

    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

    RequirementEvidenceStatus
    Implement JWT rotationintegration/jwt_test.goPASS
    Prevent SQL InjectionStatic analysis of db/user_repo.goPASS
    Secure secret handlingCode review of environment loadingPARTIAL

    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-provider module.
    • 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

    1. Scope Baseline: The agent inspects the diff, repository status, and manifests to understand the change set.
    2. Matrix Construction: It converts requirements into a matrix of verifiable claims and negative test cases.
    3. Check Execution: It runs unit, integration, and static analysis checks using the project's own toolchain.
    4. 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.

    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. 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. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    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

    Permissions

    Read Files
    Listed16 days ago

    What's inside

    Frequently Asked Questions

    While generic AI often suffers from confirmation bias, this skill operates as a skeptical auditor that ignores claims and requires 'proof of execution,' such as exit codes and test logs, before verifying code.

    The agent is language-agnostic; it automatically scans your repository's manifest files (like package.json or requirements.txt) to identify and utilize the specific testing frameworks and linters already configured in your project.

    This skill is designed to act as a 'Supervisory Agent' that can be integrated into multi-agent workflows to audit the output of coding agents before their work is merged or deployed.

    The skill is a software package for your AI agent platform that includes the verification logic, reporting templates, and the automated requirement-to-evidence matrix generator.

    Yes, the agent can perform 'Negative Testing' by attempting to trigger failures for invalid inputs, ensuring that the code doesn't just work under ideal conditions but also handles errors gracefully.

    To maintain security, the agent operates within your local environment's toolchain and requires explicit permission for any network-based checks or modifications to production files.

    New in Code Quality & Review