research-writing-assistant

Pass

Audited by Gen Agent Trust Hub on Jul 19, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill downloads Miniconda installers from the official repository at repo.anaconda.com to set up a dedicated research environment. It also communicates with several well-known academic APIs, such as CrossRef (api.crossref.org), Semantic Scholar (api.semanticscholar.org), and arXiv (export.arxiv.org), to retrieve and format literature metadata.
  • [COMMAND_EXECUTION]: The skill performs command execution to manage the local development environment and automate academic tasks. This includes executing shell and PowerShell scripts for project initialization, running conda init to configure shell profiles, installing common scientific Python packages via pip, and executing LaTeX compilation commands (xelatex). Local scripts for integrity and quality checking are run with powershell -ExecutionPolicy Bypass to ensure functionality across different environment settings.
  • [SAFE]: The identified persistence mechanism (modification of shell profiles via conda init) is a standard requirement for the Conda package manager to function correctly within a user's terminal. No evidence of obfuscation, malicious prompt injection, or unauthorized data exfiltration was found. All external resources and downloaded dependencies originate from trusted organizations and well-known technology services.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 19, 2026, 02:58 PM
Security Audit — agent-trust-hub — research-writing-assistant