The Agent Skills Directory

[PROMPT_INJECTION]: Potential for indirect prompt injection via processed video content.
Ingestion points: The scripts/fetch_transcript.py script fetches transcript data from external YouTube URLs provided by the user.
Boundary markers: The skill does not define clear delimiters or use 'ignore embedded instructions' warnings when summarizing the fetched transcript, which could lead the agent to follow instructions contained within a video's captions.
Capability inventory: The skill possesses the ability to execute Python scripts via uv run and perform file-write operations within the vault to create transcript and summary notes.
Sanitization: There is no evidence of filtering or sanitization of the transcript text to remove potential instructions or malicious metadata before it is processed by the agent.
[EXTERNAL_DOWNLOADS]: The skill relies on an external library from a public registry.
Details: The scripts/fetch_transcript.py script declares a dependency on youtube-transcript-api via PEP 723 inline metadata, which is downloaded from PyPI during execution.

youtube-ingestion