roblox-rojo-workflow
Pass
Audited by Gen Agent Trust Hub on Jun 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill directs the agent to execute standard developer tools for the Roblox ecosystem, such as
rojo,wally,stylua,selene,npm,pesde,lune, andaftman. These tools are utilized for project synchronization, package management, and Luau code linting. - [EXTERNAL_DOWNLOADS]: The workflow incorporates the use of
wallyandnpm, which fetch dependencies from their respective public registries. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it is designed to parse and interpret data from untrusted external files.
- Ingestion points: Untrusted data enters the agent context through project files including
default.project.json,wally.toml,aftman.toml, andsourcemap.json(SKILL.md). - Boundary markers: The instructions lack delimiters or explicit warnings to the agent to disregard potential instructions embedded within these external data sources.
- Capability inventory: The agent is authorized to modify source code and execute various command-line tools based on the content of the processed files (SKILL.md).
- Sanitization: There is no mention of validating or sanitizing the data ingested from the configuration files before it is used to influence the agent's actions.
Audit Metadata