social-analytics

Warn

Audited by Gen Agent Trust Hub on Feb 13, 2026

Risk Level: MEDIUMNO_CODEEXTERNAL_DOWNLOADS
Full Analysis
  • NO_CODE (MEDIUM): The core logic file scripts/main.py referenced in SKILL.md is missing. Without this file, the analyzer cannot verify if the script performs unauthorized file access, exfiltrates sensitive environment variables, or contains malicious command execution patterns.
  • Indirect Prompt Injection (LOW): The skill is designed to ingest and analyze untrusted content from social media platforms.
  • Ingestion points: Profile data and posts retrieved via requests, beautifulsoup4, tweepy, and instaloader (as indicated by dependencies).
  • Boundary markers: Unknown due to missing script code.
  • Capability inventory: Network access via Python libraries; standard subprocess execution.
  • Sanitization: Unknown. Malicious instructions embedded in social media bios or posts could potentially influence the agent's summarized output or optimization suggestions.
  • EXTERNAL_DOWNLOADS (LOW): The skill utilizes external libraries like requests and platform-specific API wrappers to download data. While the libraries themselves are standard, the source of the data (social media) is untrusted.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 13, 2026, 03:25 PM