youtube-transcript-local
Pass
Audited by Gen Agent Trust Hub on Feb 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns or security risks were identified in the skill source or instructions.
- [EXTERNAL_DOWNLOADS]: The skill utilizes the well-known 'youtube-transcript-api' library to fetch video captions from YouTube servers.
- [COMMAND_EXECUTION]: Instructions provide the user with standard commands for manual installation of dependencies and local tool execution.
- [PROMPT_INJECTION]: The skill ingests and processes untrusted transcript data from YouTube which could contain adversarial instructions. 1. Ingestion points: Data is retrieved in youtube_transcript_tool.py via the fetch_entries function. 2. Boundary markers: No delimiters or explicit instructions are provided to the agent to ignore commands within the transcript. 3. Capability inventory: The tool's capabilities are restricted to text analysis (summarization, key point extraction) and do not include file system modifications or unauthorized network activity. 4. Sanitization: The script performs text processing but does not filter for potential prompt injection patterns.
Audit Metadata