The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes powerful command-line tools including kubectl and curl to manage deployments, verify production status, and perform rollbacks. These operations are essential for the skill's stated purpose of production deployment management.
[INDIRECT_PROMPT_INJECTION]: The skill contains ingestion points for external data by reading deployment logs using kubectl logs and external API responses via curl (specifically status endpoints). While these ingestion points exist, the processing is limited to specific patterns (e.g., grep -c "ERROR") or status code checks, which effectively serves as a boundary marker to prevent instructions in that data from influencing the agent's behavior.
Ingestion points: kubectl logs -l app=signing-service, curl -s $BASE/documents, and curl -s $APP_URL/health/documenso.
Boundary markers: The skill uses specific command flags and grep patterns to isolate specific data points rather than processing full output as natural language.
Capability inventory: The skill has access to kubectl (deployment management) and curl (network operations).
Sanitization: External data is primarily used for status verification and logging count rather than being interpolated back into instructions.
[CREDENTIALS_UNSAFE]: The skill manages the DOCUMENSO_API_KEY for verification purposes. It actively promotes secure practices by instructing users to avoid .env files for production and instead use dedicated secret managers like HashiCorp Vault or AWS Secrets Manager.

documenso-prod-checklist