planning-content
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The
scripts/setup.shfile contains a command to download and install theuvpackage manager fromhttps://astral.sh/uv/install.sh. Astral is a well-known service provider for high-performance Python developer tooling. - [COMMAND_EXECUTION]: The
scripts/youtube_handler.pyscript usessubprocess.run()to executeyt-dlpfor fetching video metadata. The command is implemented using an argument list and does not invoke a shell environment, which is a secure execution practice. - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from various external sources, including files, URLs, and YouTube transcripts, to generate content outlines. While this represents a surface for indirect prompt injection (where malicious instructions could be embedded in the input documents), the skill is focused on structured analysis and outline generation, and standard model guardrails apply to the output.
Audit Metadata