Problem

When actions hang or take too long to complete, jobs can run indefinitely, causing:

• Workflows that never complete
• Dependent jobs waiting forever
• CI/CD pipelines getting stuck

This was particularly problematic for:

• HTTP actions without timeout (could hang forever)
• Background shell executions (timeout is intentionally ignored)
• Actions with retry logic (each retry could hang)

Solution

Implemented step-level timeout functionality at the action execution layer:

Implementation Details

1. Added Timeout field to Step struct (step.go:24)

   • Type: Interval (supports YAML duration format like "5m", "30s")
   • Optional field with omitempty

2. Defined DefaultStepTimeout constant (step.go:14-16)

   • Default: 5 minutes
   • Applied when no timeout is specified

3. Timeout enforcement in executeSingleAction (step.go:108-139)

   • Uses context.WithTimeout for proper timeout handling
   • Action runs in goroutine with timeout monitoring
   • Returns timeout error when exceeded

Features

• Default timeout: All actions default to 5 minutes
• Custom timeout: Can be overridden per step in YAML:
  ```yaml
  steps:
  • name: Custom timeout step
    uses: http
    timeout: 30s
    ```
• Works with retry: Each retry attempt respects the timeout
• Universal coverage: Applies to all actions (http, shell, db, etc.)

Testing

Added comprehensive unit tests (step_test.go:1289-1447):

1. ✅ TestStep_executeSingleAction_DefaultTimeout - Default 5min timeout
2. ✅ TestStep_executeSingleAction_CustomTimeout - Custom timeout enforcement
3. ✅ TestStep_executeSingleAction_CompletesBeforeTimeout - Normal completion
4. ✅ TestStep_executeActionWithRetry_Timeout - Timeout with retry logic
5. ✅ TestStep_DefaultStepTimeout_Value - Constant value verification

All tests pass:

• New timeout tests: 5/5 ✅
• Existing step tests: All passing ✅
• Full package tests: All passing ✅

Example Usage

```yaml

examples/timeout-literal.yml

jobs:

• name: Test timeouts
  steps:

  • name: Default 5min timeout
    uses: http
    with:
    url: https://example.com

  • name: Custom 10s timeout
    uses: http
    timeout: 10s
    with:
    url: https://slow-api.example.com

  • name: Timeout with retry
    uses: http
    timeout: 5s
    retry:
    max_attempts: 3
    interval: 1s
    ```

Impact

This change prevents infinite execution scenarios and ensures workflows complete in a reasonable time, improving reliability for CI/CD pipelines and monitoring systems.