The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute a shell command: python "$EMULO_PY" plugin profile-path --domain design. This relies on locating and running a local script named emulo.py from the filesystem.
[DYNAMIC_EXECUTION]: The skill uses a variable path $EMULO_PY resolved at runtime to execute a Python script, which is a form of dynamic execution based on the local environment state.
[INDIRECT_PROMPT_INJECTION]: The skill has a data ingestion surface that could be exploited if the configuration files contain malicious instructions.
Ingestion points: The skill reads the full content of files at paths returned by the emulo.py script (SKILL.md, step 4).
Boundary markers: None. The instructions do not specify the use of delimiters or warnings to the agent to ignore instructions embedded within these files.
Capability inventory: The skill uses python to execute local code and has the capability to read any file path returned by that script.
Sanitization: None. The content of the profile files is applied to the agent's context without validation or filtering.

design