Investigation View: Deployment and Configuration

The Investigation View is a new feature in the v3 user interface designed to streamline forensic analysis with structured questions, findings, and optionally AI-powered assistance. This guide walks administrators through deploying the necessary container and configuring the feature.

1. Deploying the v3 User Interface Container

Timesketch is transitioning to a new user interface built with Vue3, referred to as frontend-v3. This new interface runs in a separate Docker container alongside the existing Timesketch services. To enable it, you need to activate the v3-ui profile in your Docker Compose setup.

Prerequisites: * A standard Timesketch installation using Docker Compose, as outlined in the installation guide. * If you have an existing Timesketch deployment, please ensure to collect the latest version of the files in the data folder and nginx.conf.

Steps:

Navigate to your Timesketch directory:

Open a terminal and change to the directory where your docker-compose.yml file is located (e.g., /opt/timesketch).
Stop your current Timesketch instance (if running):

bash sudo docker compose down
Update your nginx.conf

Open the nginx.conf file in your etc/ folder and uncomment the location /v3/ section.

Start Timesketch with the v3 UI profile:

When starting your Timesketch instance, add the --profile v3-ui flag to your docker compose up command. This tells Docker to also start the timesketch-web-v3 service defined in the compose file.

bash sudo docker compose --profile v3-ui up -d

Your Timesketch instance will now be running with both the standard UI and the new v3 UI container. Your Nginx service will automatically route traffic to the appropriate container, so you can access your Timesketch instance using the same URL as before.

2. Enabling the Investigation View

The Investigation View is a powerful new feature within the v3 interface that provides a structured environment for managing investigative questions, documenting findings, and generating reports. To use it, you must enable a few settings in your timesketch.conf file. We also recommend to ensure the DFIQ (Digital Forensics Investigative Questions) data is loaded correctly.

Steps:

Enable the necessary settings in timesketch.conf:
- Open your timesketch.conf file for editing. This file is typically located in etc/timesketch/ within your main Timesketch data directory.
- Find and set the following variables to True:
  
```bash

In /etc/timesketch/timesketch.conf

DFIQ_ENABLED = True ENABLE_V3_INVESTIGATION_VIEW = True ```
[optional] Load the DFIQ Template Data:

The Investigation View works better with DFIQ templates. If you haven't already, follow the guide to load the DFIQ template data. This involves replacing the default data/dfiq directory with the contents from the official DFIQ repository.
Restart Timesketch:

For the configuration changes to take effect, you must restart your Docker containers.

```bash

In your Timesketch directory (e.g., /opt/timesketch)

sudo docker compose -f /opt/timesketch/docker-compose.yml --env-file /opt/timesketch/config.env down sudo docker compose -f /opt/timesketch/docker-compose.yml --env-file /opt/timesketch/config.env --profile v3-ui up -d ```
Verification:
- Log in to your Timesketch instance.
- Navigate to any sketch.
- You should now see a new "Investigation" tab in the top of the Left Panel.

3. Connecting the LLM Answer Drafting feature

The Investigation View includes LLM text drafting features (e.g. drafting an answer based on conclusions) that can be enabled with default LLM services.

Follow the steps described for configuring LLM Features
Add your prefered LLM provider for the new llm_synthesize feature in your timesketch.conf.
Ensure the prompt file configured in PROMPT_LLM_SYNTHESIZE exists and the prompt works for your needs.

4. Connecting the AI Investigation Agent

The Investigation View includes an experimental AI mode designed to support workflows utilizing AI Log Reasoning Agents wich automate the generation of key findings and investigative questions by analyzing timeline data. This feature is powered by a dedicated AI agent service that must be configured by an administrator.

How it Works

When a user triggers the "Generate Initial Report" in the Investigation View, Timesketch sends all timeline data to an external AI service defined in the timesketch.conf file for the log_analyzer feature. This service processes the data and sends back structured findings and questions.

IMPORTANT 1: We have developed and tested the AI feature with the experimental Sec-Gemini Log Reasoning Agent. Sec-Gemini and the Agent are not open-source, and not yet openly available. However, if you want to apply for trusted tester access, you can do this via this form: https://bit.ly/46x9GLr

IMPORTANT 2: We encurage everyone to experiment with their own AI Agent frameworks. Head to the developer section to learn more about how to design and deploy your own provider file and what Timesketch expects as response format to work with the AI Investigation View Mode.

Configuration Steps:

Configure the Agent Endpoint in timesketch.conf:

Once your agent is running, or you have gotten access to Sec-Gemini's Log Reasoning Agent, you need to tell Timesketch how to communicate with it. * Open your timesketch.conf file for editing. * Locate the LLM_PROVIDER_CONFIGS dictionary. * Within this dictionary, configure the log_analyzer section to use your custom provider endpoint or the Sec-Gemini service.

Example Configuration: ```bash

In /etc/timesketch/timesketch.conf

LLM_PROVIDER_CONFIGS = { # ... other feature configs ...
```
'log_analyzer':
{
    # This feature requires a dedicated log analyzer agent service.
    '<PROVIDER>_log_analyzer_agent': {
        'server_url': '<YOUR_AGENT_API_ENDPOINT>',
        'api_key': '<YOUR_AGENT_API_KEY>',
    }
},
```
} `` * **server_url**: This is the API endpoint URL of your deployed AI agent. * **api_key`**: If your agent requires an API key for authentication, provide it here. * Those values can vary depending on the provider you are using.
Restart Timesketch: After saving your changes to timesketch.conf, restart the Docker containers to apply the new configuration.

bash sudo docker compose -f /opt/timesketch/docker-compose.yml --env-file /opt/timesketch/config.env down sudo docker compose -f /opt/timesketch/docker-compose.yml --env-file /opt/timesketch/config.env --profile v3-ui up -d

With these steps completed, your users can now leverage the AI-powered features within the Investigation View to accelerate their forensic analysis workflows.

For Developers: Building an AI Agent

The setup of the AI log reasoning agent provider itself is a developer-focused task. Documentation on creating a compatible agent service can be found in the Developer Guides