roblox-validation-and-qa

Pass

Audited by Gen Agent Trust Hub on Aug 22, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides templates for standard local development commands including luau-lsp for type checking, selene for linting, and stylua for formatting. It also uses common shell utilities like grep, sed, find, and ls to manage project state and extract inventory information from local files. These operations are restricted to the local project environment.
  • [INDIRECT_PROMPT_INJECTION]: The skill describes a process for reading local project artifacts such as docs/BOARD.html and docs/plans/*.md to maintain a feature inventory. While this involves ingesting external data into the agent context, the data source is the project's own documentation and the process is a core functional requirement for QA tracking.
  • Ingestion points: docs/BOARD.html, docs/plans/*.md, tests/scenarios/*.md (relative project paths)
  • Boundary markers: None explicitly defined in the provided command templates for parsing documentation.
  • Capability inventory: Local shell access for project management tools and standard unix utilities; execution of Roblox Studio slash commands (/roblox-testing:test, /roblox-testing:playtest).
  • Sanitization: No specific sanitization or validation logic is defined for the content extracted from project documentation files.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 22, 2026, 04:09 AM
Security Audit — agent-trust-hub — roblox-validation-and-qa