The Agent Skills Directory

[COMMAND_EXECUTION]: The skill directs the use of the dxs CLI utility to manage remote configuration files. Specific commands include dxs configuration validate for pre-flight checks, dxs configuration upsert for updates, and dxs configuration get for retrieval. The jq utility is also used for extracting JSON payloads from platform envelopes during the editing lifecycle.
[DYNAMIC_EXECUTION]: The skill guides the generation of JSON configuration files (configurationTypeId 5) that contain TypeScript snippets within executeCodeConfig fields. These snippets are designed for runtime execution by the Datex Studio platform to drive form-level logic, such as field initialization, validation rules, and confirmation handlers.
[INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it transforms user-provided requirement briefs into executable configuration logic.
Ingestion points: The requirements brief residing in the conversation context, which serves as the source for form fields and logic (referenced in Phase 1 of SKILL.md).
Boundary markers: Explicit delimiters for untrusted requirement data are not specified in the skill's instructions.
Capability inventory: The skill performs configuration writes to a remote platform using dxs (SKILL.md) and creates logic flows that interact with form state and shell methods (references/forms.md).
Sanitization: Security is managed via the mandated dxs configuration validate command, which audits the generated JSON structure and logic before it is committed to the platform.

form-creator