prism-stack
Pass
Audited by Gen Agent Trust Hub on Jul 1, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the
Bashtool to execute framework detection viagrepand a search query via theprism_agent.uiuxPython module. - [PROMPT_INJECTION]: An indirect prompt injection surface exists where the skill processes data from untrusted project files (
package.json). - Ingestion points: Framework names and versions are extracted from project files using
grepinSKILL.md. - Boundary markers: No markers are present to distinguish between project data and instructions, nor are there warnings to ignore instructions within processed files.
- Capability inventory: The skill utilizes the
Bashtool to run shell commands and Python modules. - Sanitization: The extracted stack information (
{stack_name}) is interpolated directly into a shell command (python3 -m prism_agent.uiux search ...) without validation or escaping, which could lead to command injection if the project file content is manipulated.
Audit Metadata