The Agent Skills Directory

[DATA_EXPOSURE_AND_EXFILTRATION]: The skill includes instructions to access sensitive diagnostic information as part of its incident response procedures.
Instructions in references/immediate-actions-by-error-type.md describe how to retrieve and decode a Kubernetes secret using kubectl get secret vercel-secrets -o jsonpath='{.data.api-key}' | base64 -d for authentication troubleshooting.
Instructions in SKILL.md include a command to export production environment variables to a local file for comparison during root cause analysis: vercel env ls > env-vars.txt.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data that could theoretically be manipulated to influence agent behavior, though the risk is mitigated by its specific scope.
Ingestion points: The skill fetches JSON status updates from vercel-status.com and retrieves application runtime logs via vercel logs (documented in SKILL.md).
Capability inventory: The skill utilizes Read, Grep, and Bash (restricted to vercel and curl commands).
Boundary markers: The skill does not employ explicit delimiters or instructions to ignore embedded commands within the fetched logs or status data.
Sanitization: No specific filtering or sanitization of the external log content or API responses is documented.
[EXTERNAL_DOWNLOADS]: The skill fetches data from official platform services to provide real-time status updates.
Fetches incident and component status summaries from Vercel's official status page at vercel-status.com.
Interacts with Vercel's deployment API at api.vercel.com to manage project rollbacks and metadata inspection.

vercel-incident-runbook