Vsevolod "rapucha" Nechaev
Community@rapucha · Prague, Czech Republic
Read-only review skills for Codex that gate code changes, GitHub issues, PR scope, and sensitive data before publication.
Agent Skills by Vsevolod "rapucha" Nechaev
Showing 4 vetted skills indexed across 1 GitHub repositories.
rapuchasecond-agent-review
Run an independent read-only review of staged code changes by a second agent.
rapuchaissue-design-review
Reviews draft GitHub issues for scope authority, hidden dependencies, and acceptance criteria before creation.
rapuchasensitive-information-review
Inspect outgoing commits, PR text, and document attachments for secrets and personal data before publication.
rapuchaimplementation-scope-review
Reviews planned changes before implementation and returns a single-PR, split-required, or revise verdict.
Frequently Asked Questions About Vsevolod "rapucha" Nechaev
FAQPage SchemaWhat tasks can I perform using rapucha's skills?▼
You can run four read-only reviews: a second-agent code review of staged or branch changes, a design review of draft GitHub issues, a pre-publication scan for secrets and PII across code and documents, and a pre-implementation scope review that returns a single-PR, split-required, or revise verdict.
Who should use these review skills?▼
Engineers using Codex for nontrivial code changes, public API changes, provider integrations, refactors, or PR-ready issue branches benefit most. They also suit teams whose project policy requires validating issue premises, acceptance criteria, PR-size gates, or privacy scrutiny before publishing.
When in the development process should each skill run?▼
Run issue-design-review before an issue becomes implementation authority, implementation-scope-review before editing files, second-agent-review after implementation but before finalizing a task or PR, and sensitive-information-review before any agent-authored push or nontrivial PR creation.
Do these skills modify my code or repository?▼
No. All four skills are explicitly read-only. They inspect staged changes, draft issues, documents, and plans, then return findings or verdicts such as split-required or revise, without mutating source files, Git history, or pull requests.
What content can the sensitive-information-review skill inspect?▼
It reviews source code, Git history, unpublished PR text, PDFs, office documents, images, archives, and other attachments for secrets, credentials, unexpected PII, and sensitive content before publication or pre-commit checks.