What problem does it solve? Reviewing 1C:Enterprise BSL code manually is slow and inconsistent, especially across large extensions and change requests. This Skill automates the full review cycle: it resolves the review scope, collects evidence (linter diagnostics, naming and comment hygiene signals, project whitelists), delegates the verdicts to a dedicated code-reviewer agent, saves a structured report, and optionally routes findings to a writer agent for fixes with re-review. ## Core Features & Use Cases - Smart scope resolution: review a module, directory, extension, change request (ZN), or uncommitted git diff, with diff-focused boundaries that limit findings to changed procedures. - Evidence-driven review: aggregates linter signals, identifier hygiene (Latin characters in identifiers), comment hygiene, whitelist and mandatory-control checks before delegating to the reviewer agent. - Fix loop with re-review: after user confirmation, passes MUST_FIX findings to a code-writer agent and re-reviews (max 2 iterations); supports light review for trivial diffs and a prerelease mode for full-extension audits. - Use Case: You finished a change request touching several .bsl modules in an extension. Run the review to get a deduplicated findings report saved under the change's reports folder, then let the writer agent fix the critical findings and verify them automatically. ## Quick Start Ask the assistant to review the current BSL module or the uncommitted .bsl changes in the working tree and save the findings report.
