The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the user to install additional functionality from an external GitHub repository (ZeroPointRepo/youtube-skills) using the npx skills add command. This pattern involves downloading and integrating third-party instructions and potentially executable code from an unverified source.
[EXTERNAL_DOWNLOADS]: The skill documentation recommends adding a Model Context Protocol (MCP) server from a remote URL (https://transcriptapi.com/mcp), which introduces a dependency on an external, third-party managed service endpoint.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest YouTube transcripts, which are untrusted external data sources that could contain hidden instructions or malicious payloads intended to influence the agent's behavior.
Ingestion points: Transcript content is fetched via the /youtube/transcript endpoint of transcriptapi.com in SKILL.md.
Boundary markers: The skill does not provide instructions for using delimiters or boundary markers to separate the fetched transcript data from the agent's system instructions.
Capability inventory: The skill utilizes network operations (curl) and supports the installation of external toolsets via npx and mcp commands.
Sanitization: There are no requirements or methods specified for sanitizing, filtering, or validating the transcript text before it is processed by the agent.

youtube-research-automation