Yurii Anichkin
Community@real-case · Ukraine
From UI to AI and back again
Agent Skills by Yurii Anichkin
Showing 44 vetted skills indexed across 1 GitHub repositories.
real-caseadr-audit
Lints the Architecture Decision Record corpus for broken references, numbering issues, and invalid statuses.
real-caseaudit-run
Executes one audit from the A-01 to A-22 programme and releases reproducible Markdown and PDF reports.
real-casecommit
Generates Conventional Commits messages from staged git changes with sensitive-file detection.
real-caseadr-sync
Regenerates the architecture decisions digest in CLAUDE.md from accepted ADR records.
real-casesec-compliance
Audits code against OWASP ASVS controls and produces an evidence-based compliance gap report.
real-casetask-audit
Lints the task spec corpus for duplicate numbers, slug collisions, and dangling dependencies.
real-caseadr
Drafts structured Architecture Decision Records in MADR or Nygard format.
real-caseexplain
Explains code behavior, architecture decisions, and design rationale through structured walkthroughs.
real-casesec-scan
Audits a codebase against OWASP Top 10:2025 and produces a consolidated security report.
real-casereport-export
Export generated .marvin/ reports to PDF-ready HTML or Markdown digests using a print-quality template.
real-casechangelog
Generate structured changelogs from git commit history between tags, dates, or refs.
real-casehandoff
Captures session context into a durable handoff document for continuing work in a fresh session.
real-caseadr-review
Reviews a proposed Architecture Decision Record for structure, grounding, and readiness before acceptance.
real-casepr-merge
Merge a GitHub pull request via gh and return to the updated base branch.
real-casesec-threat-model
Generate STRIDE-based threat models with data flows, trust boundaries, and risk-scored mitigations.
real-caserefactor-audit
Audits whole-project structure to produce an evidence-backed refactoring findings register.
real-caseonboard
Guides a consent-gated first session that configures marvin in the current repository.
real-casereadme
Generate or update README.md files based on actual codebase analysis.
real-casesec-secrets
Detect leaked secrets, credentials, and private keys across source code and git history.
real-casepr-resolve
Resolve unresolved GitHub pull request review threads by applying fixes, replying, and marking them resolved.
real-casesec-iac
Reviews Terraform, Kubernetes, Docker, and CloudFormation configurations for security misconfigurations and compliance gaps.
real-caserefactor-plan
Converts refactoring findings into a sequenced, risk-annotated plan of small reversible steps.
real-casemigration-plan
Generates phased migration and refactoring plans with dependency analysis, risk inventory, and rollback strategy.
real-casesec-pentest
Generates an application-specific penetration testing checklist mapped to OWASP Testing Guide categories.
Frequently Asked Questions About Yurii Anichkin
FAQPage SchemaWhat tasks can I accomplish with Yurii Anichkin's marvin skills?▼
The 44 skills cover the full delivery cycle: drafting and ratifying ADRs, starting specced tasks, implementing and verifying them with test/lint/type-check gates, committing with Conventional Commits, creating and merging PRs, running OWASP security scans and STRIDE threat models, auditing dependencies for CVEs, planning refactorings and migrations, and exporting reports to PDF.
Who are these skills designed for?▼
Software engineers, tech leads, and security reviewers working in git-based repositories. The audit programme (A-01 to A-22) targets due-diligence and technical-assessment contexts, while sec-* skills serve teams preparing for production releases, compliance reviews against ASVS/SOC2/ISO, or penetration-testing engagements.
How do the security skills work in practice?▼
sec-scan orchestrates secrets, dependency, and IaC scans plus deep static analysis aligned to OWASP Top 10:2025. sec-gate checks only the staged diff pre-commit, sec-compliance maps code to ASVS L1/L2/L3 controls, and sec-fix turns any scanner finding (Semgrep, Snyk, Trivy, Gitleaks) into a minimal verified patch.
Do the skills modify my repository automatically?▼
Mostly no. Audit, lint, and review skills (adr-audit, refactor-audit, sec-scan, task-audit) are strictly read-only and produce reports. Mutating steps like commit, refactor-apply, and adr-accept require explicit user confirmation, and several ratification skills are human-run only with model invocation disabled.
What languages and stacks are supported?▼
task-verify auto-detects Go, Rust, Python, TypeScript, Java/Kotlin, C#/.NET, Swift, Ruby, PHP, and C/C++, with a declared-command, npm-script, or Makefile fallback for any other stack. sec-deps audits package.json, requirements.txt, Gemfile, go.mod, Cargo.toml, and pom.xml; sec-iac covers Terraform, CloudFormation, Pulumi, Kubernetes, Helm, and Dockerfiles.