choosing-a-stack

Pass

Audited by Gen Agent Trust Hub on Nov 16, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill retrieves the approved technology menu from the vendor's official repository on GitHub (Observerstudio/observer-sdlc).\n- [COMMAND_EXECUTION]: Uses CLI tools including gh api to fetch remote documentation and npm view to retrieve the latest stable version numbers for technology candidates.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted content from project files and the external menu, creating a potential surface for indirect prompt injection.\n
  • Ingestion points: Reads project-specific data (briefs, tickets) and the external markdown menu file to inform its recommendations.\n
  • Boundary markers: Utilizes a rigid "rounds" structure for the interview process which provides a framework to separate data from the agent's core instructions.\n
  • Capability inventory: Performs file system operations (read and write) and network-based CLI commands.\n
  • Sanitization: Does not include explicit sanitization or filtering logic for external data sources before processing them.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 16, 2026, 02:28 AM