split-hero
Pass
Audited by Gen Agent Trust Hub on Jul 6, 2026
Risk Level: SAFE
Full Analysis
- [DATA_EXPOSURE]: The skill accesses the credential file located at
~/.build-host/credentials.json. This is consistent with its stated purpose of interacting with the build.host platform and is considered a normal vendor-specific operation. - [DATA_EXFILTRATION]: The skill performs network operations via the build.host API (
POST /api/projects/:uuid/deploy) to facilitate project deployments. These operations target the vendor's official infrastructure and are governed by the user's existing credentials. - [PROMPT_INJECTION]: The skill includes robust safety instructions that explicitly direct the agent never to print API keys, OAuth tokens, database URLs, or webhook secrets, and to avoid asking users to input secrets in chat logs or source code.
- [COMMAND_EXECUTION]: The skill performs local repository edits to apply UI styles. It includes failure handling procedures to ensure local design contracts are preserved and accessibility standards are met.
Audit Metadata