Universal sprint orchestrator for AI agents in Claude Code. Applies Scrum-like discipline to any development task: context investigation, task decomposition with AC/DoR/DoD, hard execution gates, internal SE review, optional Codex external review, HTML reports with design system integration, LSP-based code analysis, issue tracker MCP integration. Works as platform - any plugin can declare sprint_required and ai-scrum orchestrates execution. v0.1.2: integration with claude-solo-factory kit (delegate to kit skills if available - tdd, code-review, verify, prd-writer, adr-writer, wiki-workflow).
Based on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Cancel active sprint cycle with mandatory reason. Generates partial retrospective. Cannot be undone.
Mark a step as done in active sprint. checkin-validator verifies honesty by cross-checking against tool history. Required for execution discipline - hooks may block work if no recent checkin.
Attempt to close current sprint. Triggers DoD validation gate, internal SE review, external Codex review (if available), retrospective. HARD GATE - blocks if DoD fails.
Interactive wizard to customize project DoD. Adds custom criteria to default DoD. Cannot remove mandatory baked-in criteria. Saves to .sprints/config.yaml.
Mark a step as failed with mandatory reason. Fallback can be specified via /sprint:fallback.
Bridge to codex-plugin-cc для external review. Sanitizes context, invokes Codex CLI с read-only sandbox, parses results. Falls back to DeepSeek via OpenRouter if Codex unavailable and fallback enabled.
Independent reviewer of task decomposition (fresh context, не biased). Reviews decomposition for atomicity, completeness, dependencies, AC quality. Returns APPROVED or REVISE с specific issues.
Subagent (fresh context) that updates and validates HANDOFF.md. Ensures sprint outcomes are properly captured. Detects external edits and proposes merge instead of overwrite. Required as part of sprint DoD.
Subagent (fresh context) that writes honest retrospectives. Reduces bias toward defending own work. Cross-checks numbers against execution log, flags vague language, ensures action items are specific. Used at sprint end and task cycle end.
Internal Senior Engineer reviewer subagent (fresh context). Reviews sprint outcomes across Architecture, Code Quality, Tests (extended for fake tests), Performance dimensions. Uses provided SE prompt. Required as part of DoD.
Validates Acceptance Criteria quality (specific, testable, user-focused) and verifies AC fulfillment after task completion. Used by dor-validator (quality check) and sprint-executor (verification before task close).
Validates honesty of /sprint:checkin claims by cross-checking against tool history. Detects fake claims (claiming step done without evidence), fake tests (assertions don't match test cases), and circular mocks. Used after each /sprint:checkin invocation.
External review via Codex CLI (cross-model second opinion). Activated if codex-plugin-cc installed. Falls back gracefully if Codex unavailable. Provides VERDICT: AGREE / DISAGREE / FOUND_NEW_ISSUES.
Investigates project context before task decomposition. Reads HANDOFF.md, project context briefs, glossary, README, recent commits, project structure to understand domain and patterns. Used by sprint-orchestrator at start of every sprint cycle.
Detects project design system for HTML report styling. Searches by convention paths, falls back to neutral default theme. Returns design tokens (colors, typography, spacing) for html-report-generator.
Uses power tools
Uses Bash, Write, or Edit tools
Marketplace плагинов Claude Code для AI-driven разработки.
| Plugin | Что делает | Версия |
|---|---|---|
| sdlc-audit | Глубокий разовый SDLC-аудит проекта от архитектуры до бизнес-логики. Поддерживает Python, JS/TS, Go. | 0.1.0 |
В Claude Code:
/plugin marketplace add shakhovskiya-create/shakhoff-claude-marketplace
/plugin install sdlc-audit@shakhoff-claude-marketplace
/reload-plugins
Полный аудит проекта:
/audit:full --depth=standard
Точечный аудит:
/audit:architecture
/audit:domain
/audit:code
/audit:tests
/audit:security
/audit:docs
/audit:deploy
Применение фиксов:
/audit:fix next
/audit:fix CRITICAL-001
/audit:fix --severity=BLOCKER
Подробная документация в plugins/sdlc-audit/README.md.
shakhoff-claude-marketplace/
├── .claude-plugin/
│ └── marketplace.json # манифест маркетплейса
├── plugins/
│ └── sdlc-audit/ # плагин полного SDLC аудита
│ ├── .claude-plugin/plugin.json
│ ├── commands/ # 11 команд
│ ├── skills/ # 14 skills
│ ├── agents/ # 8 subagents
│ ├── templates/ # шаблоны отчётов
│ ├── README.md
│ └── INSTALL.md
├── .claude/ # audit config для самого репо
├── README.md # этот файл
├── CONTRIBUTING.md # как добавлять новые плагины
├── LICENSE
└── .gitignore
Если нужно тестировать плагин до публикации на GitHub:
git clone https://github.com/shakhovskiya-create/shakhoff-claude-marketplace
cd shakhoff-claude-marketplace
# В Claude Code в любом тестовом проекте
/plugin marketplace add file:///absolute/path/to/shakhoff-claude-marketplace
/plugin install sdlc-audit@shakhoff-claude-marketplace
См. CONTRIBUTING.md.
Marketplace построен вокруг идеи плагин = набор инструментов для одной задачи:
MIT - см. LICENSE.
Issues и feature requests - GitHub Issues этого репо.
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimnpx claudepluginhub shakhovskiya-create/shakhoff-claude-marketplace --plugin ai-scrumDeep one-shot SDLC audit covering architecture, domain logic, code quality, tests, security, documentation, and deployment. Produces actionable backlog with severity classification and ready-to-paste fix prompts. Multi-language support: Python, JavaScript/TypeScript, Go. Integrates with ai-scrum plugin for execution discipline (sprint planning, AC/DoR/DoD gates, internal SE review).
Complete SDLC kit for solo developers using Claude Code. Six-phase pipeline (Discovery → Requirements → Architecture → Implementation → Verification → Ship) with TDD methodology, multi-perspective code review, deep verification, ADR/PRD writing, DDD modeling, design system management, autonomous overnight loops, and Karpathy code discipline rules. Was previously distributed as kit (manual copy) - now first-class plugin. Complements ai-scrum plugin (which adds hard discipline hooks and sprint state machine). Based on philosophy: plan first, verify always, recommend after every phase.
Comprehensive skill pack with 66 specialized skills for full-stack developers: 12 language experts (Python, TypeScript, Go, Rust, C++, Swift, Kotlin, C#, PHP, Java, SQL, JavaScript), 10 backend frameworks, 6 frontend/mobile, plus infrastructure, DevOps, security, and testing. Features progressive disclosure architecture for 50% faster loading.
Upstash Context7 MCP server for up-to-date documentation lookup. Pull version-specific documentation and code examples directly from source repositories into your LLM context.
Harness-native ECC operator layer - 67 agents, 271 skills, 92 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses
Develop, test, build, and deploy Godot 4.x games with Claude Code. Includes GdUnit4 testing, web/desktop exports, CI/CD pipelines, and deployment to Vercel/GitHub Pages/itch.io.
Tools to maintain and improve CLAUDE.md files - audit quality, capture session learnings, and keep project memory current.
Comprehensive PR review agents specializing in comments, tests, error handling, type design, code quality, and code simplification