Getting Started

Eptalights simplifies large-scale code analysis by transforming application artifacts into searchable datasets and APIs. The workflow consists of three main steps:

  1. Extract analysis data from your application

  2. Configure your workspace and create a build

  3. Explore the resulting analysis database

1. Extract Analysis Data from Your Application

Before running an analysis, generate the intermediate representation (IR), bytecode, or instruction data required by the platform.

Choose the extractor that matches your technology stack:

  • C/C++ Applications — Generate GIMPLE IR directly from your source code using the GCC plugin:

    eptalights-extractor-cxx

  • PHP Applications — Extract PHP bytecode for analysis using:

    eptalights-extractor-php

  • JVM-Based Applications — Extract bytecode and metadata from Java, Kotlin, Scala, Groovy, Clojure, and other JVM languages. Supports class files, JARs, WARs, EARs, and packaged applications:

    eptalights-extractor-java

  • Additional language extractors — Coming soon.

Follow the documentation for your selected extractor to generate the required analysis artifacts and store them in a directory of your choice.

2. Configure Your Workspace and Create a Build

After generating the extractor output, create a dedicated workspace for your Eptalights project. This workspace will contain your configuration file and any generated analysis databases.

Important: This step requires a valid project_id.

  • Projects can be created at: https://platform.eptalights.com/

  • The code type of your project must match the extractor used in Step 1.

  • Once your project is created, copy the project_id into your eptalights.toml configuration file.

  • For shared projects, copy the shared project ID into your configuration file instead.

Create and enter a project directory:

mkdir your-project-name
cd your-project-name

Install the Eptalights CLI and SDK:

pip install eptalights

Configure authentication:

export EPTALIGHTS_API_KEY="your-api-token"

Create an eptalights.toml file in the workspace:

project_id = "your-project-id"
extractor_output_path = "path-to-extractor-output"

Where:

  • project_id — Your Eptalights project identifier (from the platform or shared project).

  • extractor_output_path — Path to the directory containing the IR or bytecode generated in Step 1.

Create a build:

eptalights_builder --name your-unique-build-name

Notes:

  • Build names must be unique within a project.

  • Common choices: version numbers, release tags, commit hashes, CI/CD build identifiers.

After the build completes, download the generated analysis database:

eptalights_downloader --name your-unique-build-name

Important

Both eptalights_builder and eptalights_downloader must be executed from the directory containing the eptalights.toml file.

3. Explore Your Analysis Database

Once the database has been downloaded, you can query it locally using the Eptalights API.

import eptalights

api = eptalights.LocalAPI("eptalights.toml")

for fn in api.search_functions():
    print(fn.name)

Example output:

main
main
main
addNumbers
addNumbers
main
main
main
main

The local API provides access to rich analysis data, including:

  • Functions

  • Control Flow Graphs (CFGs)

  • Call Sites

  • Variables

  • File Metadata

  • Class Metadata

  • Additional program structures

Continue to the following sections to learn how to search, navigate, and analyze your application using the Eptalights API.

Questions or need help?

Join our Discord for support and discussions:

Eptalights Discord