The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/detect-model-changes.py uses the Python subprocess module to execute git and alembic commands. These calls are used to introspect the current project state (staged files, current migration revision) to assist the user in generating database migrations. The commands are constructed using argument lists, which follows security best practices for subprocess execution.
[COMMAND_EXECUTION]: The file scripts/create-migration.md utilizes dynamic context injection syntax (!command) to run shell commands such as alembic current, git diff, and date when the skill is loaded. These commands are used to gather contextual metadata (current revision, recent model changes) for the agent's migration generation task. The user-supplied $ARGUMENTS variable is used in descriptive text but is not interpolated into the shell commands themselves, mitigating command injection risks.
[DATA_EXFILTRATION]: No exfiltration patterns were found. Files containing database connection strings (e.g., references/alembic-advanced.md, rules/migration-rollback.md) use placeholder values such as postgresql://user:pass@localhost/main for demonstration purposes.

database-patterns