social-content-parser
Pass
Audited by Gen Agent Trust Hub on Jun 23, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSDATA_EXFILTRATIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill interacts with external API providers including api.bugpk.com, www.devtool.top, uapis.cn, and api.fxtwitter.com to retrieve social media metadata.
- [EXTERNAL_DOWNLOADS]: Downloads images, videos, and avatars from social media platforms and API providers to the local 'social-notes' directory.
- [DATA_EXFILTRATION]: Transmits user-provided URLs to third-party services for parsing. This activity is required for the skill's functionality and does not involve the exfiltration of sensitive local data or credentials.
- [COMMAND_EXECUTION]: Provides Node.js and shell scripts (parse_social.mjs and parse_social.sh) to automate the parsing and downloading process.
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its processing of untrusted content from social media platforms. 1. Ingestion points: Metadata such as post titles, descriptions, and author information retrieved from external APIs. 2. Boundary markers: The generated report.md and JSON files do not employ delimiters to isolate untrusted content from agent instructions. 3. Capability inventory: The skill has the ability to perform network requests (fetch) and write to the local file system (fs). 4. Sanitization: While the script cleans tracking parameters from URLs, it does not sanitize post descriptions or titles for potentially malicious instructions.
Audit Metadata