synapse-plugin-helper:test Command Workflow
Codex Adaptation
- Treat the user request or explicitly supplied text as the command arguments.
- Use the available Codex file, search, terminal, and clarification capabilities that match the workflow.
- Do not depend on Claude-only slash command variables or tool names.
Test Synapse Plugin Action
Run a plugin action in local mode for testing and development.
Arguments: the user request or explicit arguments
Workflow
Step 1: Validate Environment
Prerequisites check:
# Check synapse CLI
synapse --version 2>/dev/null || echo "ERROR: synapse-sdk not installed"
If not installed, guide user:
synapse-sdk가 설치되어 있지 않습니다.
설치: uv pip install synapse-sdk (또는 pip install synapse-sdk)
Project check:
- Check current directory has
config.yamlorsynapse.yaml - Verify the action exists in the config
- Check dependencies are installed
Step 2: Parse Arguments
$1- Action name (required)--params- Parameters as JSON string--mode- Execution mode:local(default),task,job, orremote
If action name not provided:
Which action do you want to test? Available actions:
[List from config.yaml]
Step 3: Execute Test
Run the plugin action using Synapse CLI:
synapse plugin run [action-name] --mode local --params '{"key": "value"}'
Execution modes:
| Mode | Description | Use Case |
|---|---|---|
local |
Direct Python execution | Development, debugging |
task |
Ray task execution | Integration testing |
job |
Full job execution | Pre-deployment validation |
remote |
Synapse backend execution | Requires synapse login + agent |
Step 4: Display Results
Show:
- Execution status (success/failure)
- Output from the action
- Any logged messages or metrics
- Execution time
Step 5: Handle Errors
If action fails:
- Display the error message
- Show relevant stack trace
- Suggest running
/synapse-plugin:debugfor detailed analysis - If config is out of sync, suggest
/synapse-plugin:update-config
Examples
# Test train action with default params
/synapse-plugin:test train
# Test with specific parameters
/synapse-plugin:test train --params '{"epochs": 10, "batch_size": 32}'
# Test in task mode (uses Ray)
/synapse-plugin:test inference --mode task
# Test in remote mode (requires login + agent)
/synapse-plugin:test deploy --mode remote --params '{"version": "1.0.0"}'
Quick Troubleshooting
- ModuleNotFoundError: Check
requirements.txtand install dependencies - EntrypointError: Verify entrypoint path in config.yaml matches actual file
- ValidationError: Check parameter types match Pydantic schema