Skip to main content

Application Guide

This guide explains how to create and manage Applications in Javelin, generate and use API keys, apply application-level policies, review request logs in Chronicles.

Creating an Application​

  1. Navigate to the Applications page within your gateway.
  2. From the sidebar, open Configuration.
  3. Click Create New Application button.
  4. Enter the following details in the Details tab:
    • Name – Unique identifier for your application.
    • Type – Application type (Assistant, Agent, Custom).
    • Description – A human-readable purpose for your application.
  5. Click Save.

Application Details Tab

Once Application is created:

  • A Developer Key is automatically generated.
  • This is a one time visible key, right after creation. Immediately copy and store it somewhere safe, as later you’ll only see a masked version of it.
  • This key needs to be passed in the X-Javelin-Apikey header for any route request.

Application Details Tab

Developer Tab​

The Developer tab allows you to:

  • View all keys of the application.
  • Generate new keys by clicking Add Key.
  • Delete any API key in-case it is compromised.

Usage:
For any request made through a route, include your Application API key in the header:

-H "X-Javelin-Apikey: <APPLICATION_API_KEY>"

This key authorizes requests and ensures activity is logged under your application.

Developer Tab - Add Key Modal

Application Graph Tab​

  • Displays all LLM models accessed via this application's key.
  • Shows model name, number of requests, and Last Used in a graph format.
  • Helps you analyze usage patterns and identify high-traffic models.

Application Graph Tab

Redteam Assessment Tab​

  • Displays any red team security assessments performed against this application.
  • Allows tracking of vulnerabilities, model behavior checks, and evaluation results.
  • For complete overview, see Javelin RedTeam.

Redteam Assessment Tab

Policy Tab​

  • Displays all Application-level policies currently applied to this application.
  • Policies help safeguard your LLM usage by applying filters such as content filtering, prompt injection detection, language detection, and other guardrail mechanisms.
  • For a complete breakdown of available policies and configuration options, see the Policy Configuration Guide.

Policy Tab

Chronicles Tab​

  • Displays all historical requests and responses for the application.
  • Shows:
    • Model request details
    • Input and output from the LLM
    • Guardrails data
    • Raw JSON payloads
  • Useful for debugging, auditing, and compliance checks.

Chronicles Tab

tip

Keep your Application API key safe. Do not expose it to repositories or include it in browser-delivered code. Use secure platforms to share it with the others.