The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/detect-model-changes.py executes local CLI commands to analyze project metadata and identify schema changes.
Evidence: Uses subprocess.run with list-based arguments to call git diff, git ls-files, and alembic current.
[DYNAMIC_CONTEXT_INJECTION]: The scripts/create-migration.md file uses dynamic execution placeholders to automatically populate migration files with project context.
Evidence: Employs the ! syntax to run alembic current, git diff, and system date commands at skill load time to provide environment-aware templates.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect injection as it parses local source code and incorporates user-provided arguments into migration templates.
Ingestion points: The skill reads local Python files via scripts/detect-model-changes.py.
Boundary markers: Absent; the skill lacks specific delimiters to isolate potentially malicious patterns in the ingested code.
Capability inventory: The skill can execute shell commands via subprocess and dynamic context tags.
Sanitization: Absent; no specific escaping or validation is performed on the parsed code contents.
[EXTERNAL_DOWNLOADS]: The documentation recommends the installation of third-party database management tools from a well-known service.
Evidence: rules/migration-rollback.md suggests installing the pgroll tool via Homebrew (brew install xataio/pgroll/pgroll).

database-patterns