Skip to main content

Command Execution

ValorIDE is designed to run real project commands without making the IDE feel frozen. Build, test, package, and install commands can run for minutes; the command runner keeps the user informed while protecting the agent loop from stalled output prompts.

Runtime Contract​

ValorIDE command execution follows five rules:

  • Command-output prompts time out quickly so command streams can continue when the user is not actively responding.
  • Long-running commands emit periodic progress messages with elapsed time.
  • Node execution paths have an explicit timeout for runaway commands.
  • Output streaming is fire-and-forget; a failed UI update does not block the child process.
  • Each process records start time so progress reporting is based on elapsed runtime, not guesswork.

User Experience​

When a command is still running, ValorIDE reports progress instead of waiting silently:

[Still running for 15s...]
[Still running for 20s...]
[Still running for 25s...]

This makes commands such as npm run build, cargo build, and mvn clean install suitable for agent-driven workflows where output arrives in bursts.

Operational Guidance​

Use ValorIDE command execution for:

  • project builds and test suites
  • package installs and dependency checks
  • code generation commands
  • local validation scripts

Avoid using it as an unbounded background job runner. If a command is expected to run continuously, prefer a dedicated dev server, task runner, or service process and point ValorIDE at the resulting logs or endpoint.

Troubleshooting​

If a command appears stalled:

  1. Check whether progress messages are still arriving.
  2. Verify the command is not waiting for interactive input.
  3. Run the command manually in a terminal if it requires TTY-only prompts.
  4. Increase provider or tool timeouts only when the command is expected to be slow.

Related guides: