build-evals
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes standard build and test commands including
npm run buildandnpm run test:evals. These are performed within the local environment to verify the generated test files against the project's distribution build (dist/index.js). - [REMOTE_CODE_EXECUTION]: While the skill interacts with the LLM via an API for evaluation purposes, it uses official
@umbraco-cms/mcp-server-sdk/evalstools. It does not download or execute arbitrary scripts from untrusted sources. - [CREDENTIALS_SAFE]: The skill provides instructions for managing credentials via
.envfiles or environment variables (UMBRACO_CLIENT_ID,UMBRACO_CLIENT_SECRET), which is a secure and standard practice for development environments. - [DATA_EXPOSURE]: The skill reads
.discover.jsonand tool source files to generate tests. This is restricted to the local project structure and does not involve exfiltrating sensitive data to external servers.
Audit Metadata