The SpartanX Custom Agent transforms static vulnerability data into an interactive experience. Instead of reviewing fixed reports, you can query your security data in plain language, drill into specific collections, and even import results from third-party scanners for deeper analysis.
Overview of the Interface
The Custom Agent is accessible from the left-hand navigation under "Start Custom Agent." The interface resembles a conversational AI prompt box — similar to ChatGPT or Gemini — but with one critical difference: it is directly connected to your organization's vulnerability data inside SpartanX.
The panel includes pre-built prompt categories to help you get started quickly:
Risk Management — e.g., "Show me the overall risk score for all assets and prioritize the top ten."
Deep Research — investigate specific attack paths or novel exploits
Security Posture — assess your current defensive standing
Collections, Assets, and Vulnerabilities — scoped queries for specific areas
These are starter prompts. You can edit or replace them with any question relevant to your environment.
Configuring Search Options
Three toggles control how the Custom Agent processes your queries:
Web Search — enables the agent to query public vulnerability databases alongside your internal data. Disable this to restrict responses to SpartanX platform data only.
Agentic Mode — useful in virtualized or containerized environments. In this mode, the agent can re-execute proof of concept tests to verify whether a vulnerability is genuinely exploitable.
Research Mode — enables deeper reasoning for complex topics. This mode consumes more tokens but produces more thorough analysis.
Contextualizing Queries to a CollectionBy default, the Custom Agent queries all vulnerability data across your organization. If you need to scope a query to a specific environment — for example, your PCI cardholder environment or your cloud infrastructure — you can select the relevant collection before asking your question.
Once a collection is selected, all responses are scoped to that collection. This is especially useful for:Compliance reporting (PCI DSS, HIPAA, SOC 2)
Environment-specific investigations (firewalls, APIs, internal networks)
Focused risk analysis without noise from unrelated assets
Working with Imported Scan Data
The plus icon in the Custom Agent interface opens the import functionality. This allows you to upload output from third-party scanning tools — such as Tenable Nessus, Acunetix, or Qualys — and interact with those results conversationally.
When importing, you:
Select an existing import or upload a new file
Specify the scanner tool (e.g., Nessus XML)
Set a severity threshold to filter results (critical, high, or informational for all)
Optionally store the import in a new collection
Once the playbook runs and ingests the data, you can ask questions like: "Are the critical vulnerabilities in this Nessus report actually exploitable?" — turning a static PDF into an interactive security conversation.
What's Next
The Custom Agent is most powerful when your data is well-organized. The next section covers Collections — how to create them and how they structure your assets and vulnerability data for maximum clarity and control.