youtube-transcript-extractor-api-skill
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill acts as a bridge to external content (YouTube transcripts) which presents an Indirect Prompt Injection surface.
- Ingestion points: The
scripts/youtube_transcript_extractor_api.pyscript fetches video transcripts and metadata from the external BrowserAct API, which in turn scrapes content from YouTube. - Boundary markers: The script prints the extracted transcript and metadata directly to standard output without using delimiters or boundary markers to distinguish data from instructions for the agent.
- Capability inventory: While the script itself only performs network requests and prints output, the agent executing the skill typically has access to broader capabilities such as file system operations and tool execution.
- Sanitization: There is no evidence of sanitization or filtering of the transcript text to remove potential prompt injection payloads before it is passed to the agent.
- [EXTERNAL_DOWNLOADS]: The skill performs network operations to external services.
- The script
scripts/youtube_transcript_extractor_api.pycommunicates withapi.browseract.comto initiate and poll for transcript extraction tasks. This domain is managed by the skill's vendor. - The skill uses the
requestsPython library to facilitate these connections.
Audit Metadata