substack-notes-scraper

Pass

Audited by Gen Agent Trust Hub on Jun 17, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill uses WebFetch to download rendered HTML content from substack.com, which is a well-known service. This behavior is consistent with the stated purpose of scraping newsletter metrics.
  • [COMMAND_EXECUTION]: The skill instructions direct the agent to generate and execute a Python script using the openpyxl library. This is used to create a structured .xlsx file with conditional formatting and summary statistics, which is standard functionality for data processing tools.
  • [PROMPT_INJECTION]: As the skill ingests third-party content from Substack, it possesses an inherent surface for indirect prompt injection.
  • Ingestion points: Step 2 and Step 4 in SKILL.md describe fetching and parsing rendered HTML from user-provided Substack URLs.
  • Boundary markers: None explicitly defined for the parsed note content.
  • Capability inventory: Python file-write capabilities for Excel generation as defined in Step 8.
  • Sanitization: Step 4 includes instructions to strip HTML tags from extracted text.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 17, 2026, 12:10 AM
Security Audit — agent-trust-hub — substack-notes-scraper