kbo-results
Pass
Audited by Gen Agent Trust Hub on May 12, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructions specify the global installation of the
kbo-gamepackage from the npm registry. This introduces a dependency on a third-party package not maintained by a well-known vendor. - [COMMAND_EXECUTION]: The skill uses shell commands to install software (
npm install -g) and executes an inline Node.js script using thenode --input-type=moduleflag with a heredoc. - [REMOTE_CODE_EXECUTION]: Within the Node.js script, the skill performs a dynamic import (
await import(entry)) of a file located at a path constructed at runtime using theGLOBAL_NPM_ROOTenvironment variable. This pattern is used to bridge the globally installed package with the execution environment. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes untrusted data from external sports results.
- Ingestion points: Data returned from the
getGamefunction in the Node.js script withinSKILL.md. - Boundary markers: No explicit delimiters or instructions are provided to the agent to ignore potential instructions embedded in the game data.
- Capability inventory: The skill possesses the ability to execute shell commands and install packages.
- Sanitization: No data validation or escaping is applied to the retrieved game information before it is processed by the agent.
Audit Metadata