generating-custom-lightning-type
Pass
Audited by Gen Agent Trust Hub on Jul 29, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill acts as a technical reference and generation guide for Custom Lightning Type (CLT) metadata. It includes strict adherence to platform validation rules and bundle structures to ensure compatibility and security within the Salesforce ecosystem.
- [DATA_EXFILTRATION]: No patterns of credential theft, sensitive file reading, or unauthorized data transmission were found. The skill operates within designated project directories for metadata development.
- [REMOTE_CODE_EXECUTION]: The skill does not download or execute external scripts, binary payloads, or untrusted third-party dependencies. All activities involve generating static JSON configuration files.
- [COMMAND_EXECUTION]: The skill utilizes platform-specific metadata actions (e.g., discoverUiComponents) for UI discovery and schema retrieval. These are standard, scoped capabilities used according to the intended design of the agent platform.
- [PROMPT_INJECTION]: The skill provides instructions for generating structured data based on user input. While it represents an indirect prompt injection surface, it mandates strict platform validation (e.g., unevaluatedProperties: false and removal of the $schema keyword) which effectively mitigates the risk of malicious payload injection into the resulting metadata.
Audit Metadata