The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or unauthorized data access were detected in the skill files. The skill focuses on legitimate administrative tasks using official Vercel tools.
[COMMAND_EXECUTION]: The skill documents standard Vercel CLI operations such as vercel deploy, vercel env pull, and vercel logs. These commands are used for deployment management and monitoring within the user's authenticated scope.
[PROMPT_INJECTION]: The skill includes procedures for diagnosing failures by reading build and runtime logs. This constitutes an indirect prompt injection surface as logs are external data processed by the agent.
Ingestion points: vercel logs and mcp_com_vercel_ve_get_runtime_logs mentioned in mcp-reference.md.
Boundary markers: Absent.
Capability inventory: The skill documents commands to promote, redeploy, and modify environment variables.
Sanitization: Not specified; the skill assumes the user is inspecting logs in a trusted development context.

vercel