The Agent Skills Directory

Command Execution: The skill uses the vercel CLI and standard utilities like curl and tail to retrieve logs and deployment metadata. These commands are used within the context of troubleshooting application failures.\n- Data Access and Inspection: The skill analyzes runtime logs, browser console output, and environment configurations. This is a necessary step for root-cause analysis, although users should be mindful that logs may contain sensitive information depending on the application's logging configuration.\n- Indirect Prompt Injection Surface: The skill processes external data which could theoretically contain instructions. \n
Ingestion points: Reads runtime logs (vercel logs), browser console output, and workflow statuses as described in SKILL.md.\n
Boundary markers: None explicitly defined for log content within the skill instructions.\n
Capability inventory: Executes shell commands via vercel CLI, curl, and tail as defined in SKILL.md.\n
Sanitization: None specified for the content of retrieved logs before reporting to the user.

investigation-mode