skills/moshe-ship/mkhlab/safha/Gen Agent Trust Hub

safha

Pass

Audited by Gen Agent Trust Hub on Jun 26, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes the 'safha' command-line tool to perform web scraping, sitemap parsing, and text processing operations.
  • [EXTERNAL_DOWNLOADS]: The skill requires the 'safha' binary as specified in the metadata frontmatter. This is a standard environment configuration requirement.
  • [DATA_EXFILTRATION]: Includes commands to fetch data from user-provided URLs via scraping tools. This behavior is consistent with the skill's stated purpose of gathering training data and does not target sensitive local files or environment variables.
  • [PROMPT_INJECTION]: Analysis of the instructions and metadata found no evidence of attempts to override agent behavior, bypass safety protocols, or leak system prompts.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 26, 2026, 01:16 PM
Security Audit — agent-trust-hub — safha