platform-custom-field-generate

Pass

Audited by Gen Agent Trust Hub on Aug 13, 2026

Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection by ingesting untrusted data from the user and incorporating it into metadata files saved to the local environment.
  • Ingestion points: User-provided field labels, descriptions, and picklist values defined in SKILL.md.
  • Boundary markers: Absent; there are no instructions provided to wrap untrusted content in specific delimiters or to add warnings to ignore embedded instructions within the user-supplied data.
  • Capability inventory: File-write capabilities (e.g., force-app/main/default/objects/) to the local project structure as described in the SKILL.md instructions.
  • Sanitization: While the skill correctly mandates CDATA blocks for formula fields and provides naming conventions, it lacks explicit instructions to sanitize or escape special characters for other text nodes like <description> and <inlineHelpText>.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 13, 2026, 12:59 PM
Security Audit — agent-trust-hub — platform-custom-field-generate