Configure and Manage FastGPT Agent VM Lifecycle Scripts

Startup Script Editor Access

Under the Computer Configuration section of the Agent Configuration Panel, you can write standard Shell commands directly inside the Startup script (sh) code editor. !Startup Script Editor This editor accepts valid Shell syntax to define custom environment setup logic for your agent’s virtual machine.

Script Execution Sequence and Scope

When a new agent session starts or the virtual machine is reconstructed, the system executes configured scripts sequentially in the background:

1. App Startup Script: The custom Shell script saved in the Agent panel executes inside the virtual machine’s working directory, which is typically /workspace. Its core purpose is to prepare app-specific dependencies and runtime environments required for the agent to function.
2. Skill Entrypoint: If your Agent is associated with published skills, the initialization entrypoint script (e.g., entrypoint.sh) bundled within the skill package will be extracted and executed in the skill’s dedicated deployment directory immediately after the App Startup Script completes.

⚠️ Transactional Skill Deployment: During skill deployment, packages are first extracted to a temporary folder (e.g., .tmp-<versionId>-<random>). Upon successful decompression, the folder is atomically renamed to the formal version directory to prevent corrupted partial extractions that could break skill functionality.

Lifecycle Flowchart Reference

The official FastGPT documentation provides a visual lifecycle flowchart that maps the full sequence of virtual machine events tied to agent script execution. !Lifecycle Flowchart This chart illustrates triggers for script execution, including new session launches and virtual machine reconstructions, as well as the ordered execution of app and skill scripts.

Step-by-Step Script Validation Workflow

1. Navigate to the Computer Configuration subsection of the Agent Configuration Panel.
2. Input a test Shell command (e.g., echo "Startup script executed" > /workspace/startup-log.txt) into the Startup script (sh) editor.
3. Save the updated agent configuration and launch a new agent session or trigger a virtual machine reconstruction.
4. Access the virtual machine’s /workspace directory to confirm the startup-log.txt file exists and contains the test message.
5. For agents linked to skills, verify the skill entrypoint script runs by checking logs in the skill’s deployment directory.

Source: FastGPT official documentation and source