academic-research-suite

Pass

Audited by Gen Agent Trust Hub on Aug 11, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill expects core components, templates, and agent definitions to be downloaded from the author's repository (github.com/crazymsn/academic-shcolar-skills) and placed into the ars/ subdirectory.
  • [COMMAND_EXECUTION]: The documentation instructs the user to execute local installation scripts (install-codex-skill.ps1 or install-codex-skill.sh) to set up the tool suite environment.
  • [PROMPT_INJECTION]: Indirect prompt injection surface identified in the handling of user-supplied research data.
  • Ingestion points: User research questions, paper topics, and manuscript drafts provided during research and writing workflows.
  • Boundary markers: Absent; the router does not define specific delimiters or instructions to ignore embedded commands in the user-provided research material.
  • Capability inventory: Built-in web browsing for citation verification and file-system read access for loading sub-skill workflows from the ars/ directory.
  • Sanitization: Absent; no specific validation or filtering of research inputs is implemented in the routing logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 11, 2026, 01:03 AM
Security Audit — agent-trust-hub — academic-research-suite