Menu
This skill outlines the workflow for verifying the Redbox Portal web interface using the Agent's built-in browser capabilities and natural language test scripts.
Instalar com Codex ou Claude Copie este prompt, cole no Codex, Claude ou outro assistente e deixe que ele revise a página da skill e instale para você.
Baseado na classificação ocupacional SOC
ReDBox uses embedded Angular applications within EJS views instead of a single SPA. Each feature is often its own Angular project.
Angular services in ReDBox typically extend `HttpClientService` to handle base URLs, configuration, and CSRF protection.
ReDBox controllers live in `@researchdatabox/redbox-core` and are surfaced to Sails.js through generated shims.
Create detailed feature designs, implementation plans, and task lists for ReDBox/Redbox features that use Waterline models, services, webservice (REST) controllers, ajax controllers, embedded Angular apps, views, and navigation config. Use when asked to design/plan a feature or produce a task list for this stack, especially when explicit sections, non-SPA Angular patterns, and interleaved tests are required.
Review completed ReDBox feature implementations against a growing ruleset and emit structured issues for agent remediation.
Build or update ReDBox form field components end-to-end across Angular form rendering and sails-ng-common config typing/visitor infrastructure. Use when creating a new form component class/model, adding component config properties, wiring component dictionaries, handling config migration/template extraction, or adding form-component tests.
| name | Web Interface Verification |
| description | This skill outlines the workflow for verifying the Redbox Portal web interface using the Agent's built-in browser capabilities and natural language test scripts. |
This skill outlines the workflow for verifying the Redbox Portal web interface using the Agent's built-in browser capabilities and natural language test scripts.
Start the application in development mode with the mount profile. The dev:run script applies the mount profile implicitly (no extra flags required).
Important: Capture the CommandId returned by this command to monitor logs later. The CommandId is typically a UUID or numeric ID printed near startup; look for it before or around the "Sails lifted" readiness marker.
npm run dev:run
Example output (capture the value after CommandId:):
CommandId: 9f2b5c3a-1d4e-4a6a-9e5b-0a1234567890
...Sails lifted...
Extract it from the logs with a one-liner (or copy the CommandId: line directly):
grep -m 1 "CommandId:" .tmp/logs/dev-run.log
Wait for the application to be ready (look for "Sails lifted").
Use the browser_subagent tool to launch a browser session.
Follow the natural language instructions provided by the user. Example Script:
/researcher/home."Agent Responsibilities:
While the browser session is active or after a step fails, check the logs of the running application using the command_status tool and the captured CommandId.
After completing the script or encountering a failure, report the results to the user.