The Agent Skills Directory

[COMMAND_EXECUTION]: The skill facilitates the execution of local shell scripts and Python modules to manage a development stack.
Evidence: Invocations of scripts/dev-local.sh with various subcommands (up, down, logs, etc.).
Evidence: Use of uv run python -m treg to start the application server.
[EXTERNAL_DOWNLOADS]: The skill manages project dependencies using the uv package manager.
Evidence: The skill description mentions that uv sync is executed if the virtual environment (.venv) is missing.
[INDIRECT_PROMPT_INJECTION]: The skill accepts user-supplied arguments for the CLI, creating a surface for indirect prompt injection if those arguments are processed unsafely by the underlying application.
Ingestion points: scripts/dev-local.sh cli <args> (SKILL.md).
Boundary markers: None explicitly mentioned in the skill instructions.
Capability inventory: Shell command execution via script invocation.
Sanitization: Not explicitly implemented in the skill instructions; relies on the underlying shell script/application logic.
[DATA_EXPOSURE]: The skill interacts with configuration files but includes explicit measures to protect the user's primary configuration.
Evidence: The skill sets HOME=scripts/.dev-home to ensure the local development environment does not overwrite the real ~/.treg/config.json file.

dev-local