Skip to main content

Generative AI Copilot

Cyver Core is designed to streamline and automate your entire pentest operation. The GenAI Copilot takes this automation to the next level by integrating the power of generative AI directly into your reporting and findings management processes.

We understand that one of the most time-consuming parts of a penetration test is writing clear, consistent, and high-quality documentation, especially executive summaries, detailed finding descriptions, and actionable recommendations.

The GenAI Copilot is built to tackle these specific tasks, acting as your intelligent writing assistant. It helps you:

  • Automate Tedious Writing: Instantly generate full report sections or detailed descriptions based on your project's data.
  • Ensure Consistency: Maintain a standard tone, format, and quality across all reports and for all pentesters.
  • Save Valuable Time: Free your team from repetitive writing so they can focus on what matters most: complex analysis and vulnerability discovery.

You can use the GenAI Copilot in two primary ways: generating entire Report Sections and populating specific Finding Fields.

Before using these features, let's look at how the Copilot understands what to generate and how you can customize its instructions (prompts).

GenAI Copilot Settings

The text generated by the Copilot is based on templates called Prompts. You have full control to edit, create, and test these prompts to match your company's voice and reporting standards.

You can access these settings by navigating to Settings > GenAI Copilot Settings from the main menu.

This settings page is divided into two main tabs:

1. Report Prompts

This tab controls the prompts used to generate for your report sections (e.g., "Executive Summary," "Recommendations").

  • You can Edit or Delete existing prompts.
  • You can create new ones using the + New AI Prompt button.
  • When editing or creating, you define the Name (how it appears in the menu), Status (Published/Draft), and the Prompt itself.
  • The Prompt field uses tokens like {Client_Name} or {Finding_Details} to pull live project data into your instructions for the AI.

2. Finding Field Prompts

This tab controls the prompts used to fill specific finding fields: Description, Background Information, Remediation, Likelihood Description, Impact Description.

  • The key difference in this settings menu, compared to Report Prompts, is the addition of a Field Name dropdown menu.
  • This dropdown is what links your prompt to a specific field on the Edit Finding page. For example, if you set the Field Name to "Remediation," this prompt will only appear as an option when you click the GenAI Copilot button for the "Remediation" field.

Using GenAI to Generate Report Sections

This method is perfect for creating first drafts of major report components based on all the findings in your project.

  1. Navigate to your project and click the Report tab.
  2. Select the report section you wish to generate, such as "Executive Summary".

  3. Click the Editor to view the text-editing interface.

  4. In the editor toolbar, click the Generate Text button.

  5. A GenAI Copilot window will appear, presenting you with a list of all Published prompts from the Report Prompts settings.

  6. Select the prompt you want to use. The Copilot will immediately begin generating the text.

  7. Once generated, you have four options:
    • Apply: This inserts the AI-generated text directly into your report section editor.
    • Regenerate: If you're not satisfied with the output, click this to generate a new version.
    • Back: This discards the current text and takes you back to the prompt selection screen.
    • Close: This will discard the generated text and close the window.

Example: Creating and Using a New Report Prompt

Let's create a new, specific prompt for a "Technical Risk Summary" and then use it in our report.

Part 1: Create the New "Technical Risk Summary" Prompt

  1. Go to Settings > GenAI Copilot Settings.

  2. On the "Report Prompts" tab, click + New AI Prompt.

  3. Fill in the fields:
    • Name: Technical Risk Summary
    • Status: Published

    • Prompt: Write a technical risk summary for {Client_Name} based on the {Findings_Details}. Focus only on Critical and High severity findings. Do not provide recommendations, only analyze the technical impact and exploitability.

  4. (Optional) Click Test Prompt to see what it generates.
  5. Click Save.

Part 2: Use the New Prompt in a Report

  1. Go to your project's Report tab.

  2. Click on the relevant report section or create a new section. In our example, we are creating a new section via 3 dots > Manage sections > + New Section > "Technical Risk Summary".

  3. Click Editor.

  4. Click Generate Text.

  5. In the GenAI Copilot window, you will now see your new "Technical Risk Summary" prompt in the list of options.

  6. Select it, wait for generation, and click Apply. The text will be inserted.

Using GenAI to Populate Finding Fields

This method is ideal for speeding up the documentation of individual findings. The Copilot can write descriptions, impact analyses, and remediation steps for you, one field at a time.

  1. From your project, navigate to the Findings tab.

  2. Select any finding to view its details (e.g., "Code Execution via File Upload").

  3. Click the Edit button to open the Edit Finding page.

  4. You will see a GenAI Copilot button next to several text fields: Description (on the "Basic" tab), Impact Description and Likelihood Description (on the "Classification" tab), Background Information and Remediation (on the "Recommendations" tab).

  5. Click the GenAI Copilot button for the field you want to fill.

  6. Select a prompt. The AI will generate the content and present it for review.

How Prompts are Displayed: The menu that appears will only show prompts that have been assigned to that specific field in the GenAI Copilot Settings.

  • Since we had only one prompt assigned to the "Impact Description" field, we see only one option.
  • If we have created three different prompts and assigned all of them to the "Impact Description" field, we would see all three as options to choose from.

  1. Click Apply to populate the field with the generated text.

Example: Creating and Using a New Finding Field Prompt

Let's create a highly specific "Remediation for Java" prompt and use it on a finding.

Part 1: Create the New "Java Remediation" Prompt

  1. Go to Settings > GenAI Copilot Settings.
  2. Go to the Finding Field Prompts tab.

  3. Click + New AI Prompt.

  4. Fill in the fields:
    • Name: Java Specific Remediation
    • Field Name: Select Remediation from the dropdown menu 
    • Status: Published
    • Prompt: When provided with vulnerability details, you MUST propose a detailed technical remediation plan specifically for a Java and Spring Boot environment. Your response must include: 
      1. A brief explanation of *why* the provided context is vulnerable. 
      2. A secure code example using Java PreparedStatement. 
      3. An explanation of how modern ORMs like Spring Data JPA help prevent this issue. 

Finding Name: {Finding_Name} 

Vulnerability Description: {Finding_Description} 

Vulnerability Type: {Finding_Vulnerability_Type} 

Background Info: {Finding_BackgroundInformation} 

CWE(s): {Finding_CWEs}

  1. Click Save.

Part 2: Use the New Prompt on a Finding

  1. Go to a project and open any finding (e.g., "SQL Injection").

  2. Click Edit.

  3. Navigate to the Recommendations tab.

  4. Click the GenAI Copilot button located next to the Remediation field.

  5. You will now see "Java Specific Remediation" in the list of options (along with any other prompts you have assigned to the Remediation field).

  6. Select it, wait for generation, and click Apply. The detailed Java remediation steps will be inserted into the field.

  7. Click Save.

Data Privacy

In order to generate the text, we use project, client, findings and asset data from the platform. We use internal AI models (ChatGPT based), no data is shared externally with 3rd parties.

Was this article helpful?