What problem does it solve? DIVA source files (.dhsp, .dhsq, .dhsd, .dhsf, .dhsi, .dhse, .dhpt, .dhps, .sql) for Divalto Harmony must comply with roughly 100 known anti-patterns, syntax rules, dictionary conventions, and CI quality checks that were historically only caught by nightly CI pipelines. This Skill shifts those checks left, producing a severity-classified error report before compilation or after code generation. ## Core Features & Use Cases - Rule-based linting across 9 file types: Checks ~100 rules grouped by category (S syntax, L forbidden language keywords, Z zoom architecture, M module checks, F framework, R RecordSql, D dictionary, E screen masks, P project files, CI shift-left quality gates), each classified as error, warning, or info. - Structured JSON reports: Outputs findings with rule code, severity, line number, message, and code context, filterable by rule or severity, enabling a generate-validate-correct feedback loop. - Use Case: After generating a .dhsp zoom program, run the linter to catch guaranteed bugs like string concatenation with + instead of &, ForEach keywords, or OverWrittenBy pointing to a source file instead of the compiled .dhop, before compilation fails. ## Quick Start Ask the AI to lint a DIVA source file, for example: analyze the file rttzfamrglt_sql.dhsp with the linting-diva-code skill and list all errors and warnings.
