The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides numerous templates and examples for executing commands via official cloud CLI tools and database utilities to manage migrations, snapshots, and data imports.
Evidence: references/cloud-specific-migrations.md contains command patterns for aws rds, gcloud sql, and az sql CLI tools.
Evidence: references/migration-phases.md includes examples for pg_dump, mysqldump, pgloader, and ora2pg for data and schema transfer.
[INDIRECT_PROMPT_INJECTION]: The skill workflow requires the agent to ingest and analyze external data, such as database schemas, triggers, and inventory metadata, which constitutes a potential attack surface if that data contains malicious instructions.
Ingestion points: SKILL.md (Step 1: Assessment and Planning) and references/migration-phases.md (Phase 1: Assessment and Planning) instruct the agent to analyze the source database version, complexity, and schema objects.
Boundary markers: No specific delimiters or safety warnings are provided for the agent to use when processing this external metadata.
Capability inventory: The skill lists capabilities including network access (cloud CLI, database connections), file system access (creating/restoring backups), and schema modification.
Sanitization: No explicit sanitization or validation steps are defined for the ingested metadata.

database-migration