youtube-acquire
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes the
yt-insightstool using user-providedSOURCEarguments. There is a risk that unsanitized user input could lead to command injection when the agent runs the CLI tool in the main session. - [INDIRECT_PROMPT_INJECTION]: The skill ingests external data (transcripts and metadata) from YouTube which could contain malicious instructions designed to influence the agent when the corpus is later processed. 1. Ingestion points: YouTube source material acquired into the yt-insights corpus (SKILL.md). 2. Boundary markers: Absent; the skill does not instruct the agent to use delimiters or ignore embedded instructions for the acquired data. 3. Capability inventory: Shell command execution via
yt-insightsand local corpus modification (SKILL.md). 4. Sanitization: The instructions do not define any sanitization or validation steps for the content fetched from YouTube.
Audit Metadata