The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes subprocess.run in scripts/run.py, scripts/setup_environment.py, and scripts/__init__.py to automate environment setup, install dependencies, and wrap script execution. These calls are used for the primary purpose of managing the skill's internal infrastructure and isolated virtual environment.
[EXTERNAL_DOWNLOADS]: During initial setup, the skill downloads required Python packages from official registries and browser binaries (Chrome/Chromium) through the patchright automation library.
[PROMPT_INJECTION]: The skill implements a 'Follow-Up Protocol' by appending instructions to the end of answers in scripts/ask_question.py. This directs the AI agent to evaluate the completeness of the information and perform additional queries if gaps exist, acting as a control-flow mechanism for the research process.
Ingestion points: External text retrieved from the NotebookLM web interface (scripts/ask_question.py).
Boundary markers: Not present for the ingested text.
Capability inventory: Subprocess execution for script wrapping and dependency management (scripts/run.py, scripts/setup_environment.py).
Sanitization: None detected for the retrieved answers before they are returned to the agent context.

notebooklm