More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    Secure Code Vulnerability Review Engineer

    1

    Senior application security engineer for identifying, prioritizing, and fixing code-level vulnerabilities.

    $29.99

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    Secure Code Vulnerability Review Engineer

    Secure Code Vulnerability Review Engineer

    Example session with this skill installed

    Review this Node.js snippet for security flaws:

    app.get('/user', (req, res) => {
      const query = "SELECT * FROM users WHERE id = " + req.query.id;
      db.execute(query, (err, result) => { res.send(result); });
    });
    
    • Read your context and instructions
    • Compiled the secure code vulnerability

    I have completed the security review of the provided Node.js endpoint. The analysis identified a critical injection vulnerability due to unsafe string concatenation in a database query.

    Security Review Report

    Executive Summary
    The reviewed code contains a critical SQL Injection vulnerability. An unauthenticated attacker can manipulate the database query to bypass authorization or exfiltrate data. Immediate remediation is required.

    Findings

    IDTitleSeverityConfidenceStatus
    SEC-001Unparameterized SQL InjectionCriticalHighConfirmed
    • Affected Path: app.get('/user', ...)

    Evidence: The variable req.query.id is concatenated directly into the SQL string without validation or parameterization.

    Impact: Attackers can append SQL commands (e.g., 1 OR 1=1) to view all user records or use UNION select statements to dump other tables.

    • Recommended Fix: Use prepared statements with parameterized queries.
    // Recommended Remediation
    app.get('/user', (req, res) => {
      const query = "SELECT * FROM users WHERE id = ?";
      db.execute(query, [req.query.id], (err, result) => {
        if (err) return res.status(500).send("Database error");
        res.send(result);
      });
    });
    
    • Regression Test:
      1. Pass a non-numeric string like ' OR '1'='1 to the id parameter.
      2. Verify the application returns an error or no results rather than the entire user table.

    Review Gaps
    The database driver configuration (db) was not provided; I assume it supports standard parameterization. No input validation middleware was visible in the snippet.

    Next steps

    1. Implement the parameterized query patch shown above.
    2. Add a validation layer (e.g., joi or zod) to ensure id is a numeric integer.
    3. Review all other database interactions in the controller for similar concatenation patterns.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Manual security audits are slow and prone to oversight. This skill acts as a senior application security engineer to identify, validate, and document vulnerabilities in your source code or pull requests.

    What it does

    • Vulnerability identification traces untrusted input through validation, authorization, and sensitive sinks.
    • Severity rating prioritizes findings based on reachability, privileges, and blast radius.
    • Remediation guidance provides precise patches and framework-specific safe alternatives.
    • Regression testing generates specific test cases to verify fixes and prevent future reintroduction.
    • Evidence-backed reporting produces structured summaries including data flow, preconditions, and confidence levels.

    How it works

    1. Scope definition where you provide files, diffs, or repository excerpts and define the trust boundaries.
    2. Security inspection where the agent reviews code for injection, auth failures, crypto weaknesses, and logic flaws.
    3. Trace validation to confirm if a dangerous pattern is actually reachable by an attacker.
    4. Report generation providing an executive summary, prioritized findings, and an action plan.

    Frameworks & tools

    This skill is language-agnostic and works with any source code, configuration files, or pull-request diffs. It is particularly effective for web frameworks, API backends, and cloud-native applications.

    Why this beats prompting it yourself

    Generic prompts often flag "scary" functions without checking if they are reachable. This skill follows a rigorous defensive methodology that distinguishes between suspicious patterns and confirmed, exploitable vulnerabilities with clear impact analysis.

    Use cases

    • Reviewing pull requests for high-risk security regressions before merging.
    • Auditing legacy codebases for hidden injection or authorization flaws.
    • Validating the implementation of sensitive features like MFA or session handling.
    • Generating security-focused unit tests for critical business logic.

    Known limitations

    Does not execute live code or access production systems. Accuracy depends on the amount of technical context and trust boundary information provided in the snippets.

    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.

    Requires source code or diffs to be provided. Cannot perform live runtime exploitation. Accuracy is limited when critical dependencies or architectural context are missing from the input.

    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 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 10 days ago

    Permissions

    Read Files
    Write Files
    Network Access
    Listed10 days ago

    What's inside

    Frequently Asked Questions

    The skill performs a deep data-flow analysis to determine if untrusted input can reach sensitive functions, significantly reducing false positives compared to standard AI prompts that flag any suspicious-looking code.

    This skill is language-agnostic and effective across all major programming languages (Python, Java, JavaScript, Go, C#, etc.), as well as configuration files like Dockerfiles or Kubernetes manifests.

    The skill requires the relevant code snippets, diffs, or files as input; it does not require direct access to your private repositories or production environments.

    Every purchase includes the core security review logic, documentation on how to define trust boundaries for the agent, and lifetime access to any logic updates or vulnerability signature improvements.

    Yes, the agent is designed to analyze pull-request diffs, allowing you to catch security regressions and logic flaws before code is merged into your main branch.

    While the agent is highly proficient at identifying logic flaws and common vulnerabilities (OWASP Top 10), its accuracy depends on the amount of architectural context and trust boundary info provided in your prompts.

    New in Code Quality & Review