The Agent Skills Directory

[COMMAND_EXECUTION]: The instructions direct the agent to interact with the system using the odev CLI, facilitating tasks like starting/killing servers and managing databases.
[DYNAMIC_EXECUTION]: The skill details methods for executing Python code and shell commands through odev shell --script and odev venv -c respectively, which are core features for debugging and environment setup.
[EXTERNAL_DOWNLOADS]: The skill provides instructions for installing Python packages within a virtual environment using pip through the odev venv command.
[INDIRECT_PROMPT_INJECTION]: The skill facilitates the processing of variable inputs, such as database names and module paths, which are interpolated into shell commands.
Ingestion points: Command arguments in SKILL.md including <database>, <modules>, and custom scripts passed to the --script flag.
Boundary markers: None explicitly mentioned in the usage rules.
Capability inventory: File system access for database resources, process lifecycle management, and arbitrary code execution within specific development contexts.
Sanitization: Instructions focus on syntax and parameter ordering rather than input validation or escaping.

odev