By Valdemird
Stack-agnostic spec-driven workflow for Claude Code: 3 subagents (spec-writer, architect, implementer), 13 curated skills (spec-* workflow, audit suite, bugfix, frontend-design, skill-creator), 4 deterministic quality-gate hooks, and an AUTO_MODE for autonomous multi-phase implementation.
Staff Engineer subagent that validates a spec against the real codebase, detects technical risks, and produces phased TDD implementation plans. Invoke during spec audit and plan generation.
Senior Developer subagent that implements a pre-approved plan with strict TDD (Red/Green/Refactor). Stops on STOP-protocol triggers (existing code contradicts plan, missing dependency, broken existing test, ambiguous plan). Invoke during phase implementation.
Technical PM subagent that turns vague ideas into structured, actionable spec documents. Asks uncomfortable edge-case questions, writes complete specs that an architect can audit. Invoke when the user wants to create a new spec.
Run a thorough architecture review of the codebase or a specific module. Evaluates structure, separation of concerns, scalability, coupling, and alignment with best practices.
Disciplined bug-fix workflow with TDD (diagnose → failing test → minimal fix → full verification). Forces root-cause analysis before writing code, eliminating "fix → fail → re-fix" cycles. Triggers on "fix the bug", "X is broken", "not working", "regression".
Create distinctive, production-grade frontend interfaces with high design quality. Use this skill when the user asks to build web components, pages, artifacts, posters, or applications (examples include websites, landing pages, dashboards, React components, HTML/CSS layouts, or when styling/beautifying any web UI). Generates creative, polished code and UI design that avoids generic AI aesthetics.
Expert mobile interaction auditor for web apps. Use this skill whenever the user wants to audit, review, fix, or improve mobile UX — including touch interactions, virtual keyboard handling, safe areas, responsive layouts, scroll behavior, device-specific quirks, or any mobile edge case. Also trigger when the user mentions: mobile bugs, touch targets, swipe issues, viewport problems, keyboard pushing content, notch/safe-area issues, iOS Safari quirks, Android Chrome issues, responsive breakpoints, hover states on mobile, tap delay, input zoom, orientation changes, overscroll behavior, or "it works on desktop but not on phone". Even if the user doesn't say "mobile audit" explicitly, use this skill whenever the problem is clearly mobile-specific.
Run a thorough performance review on the codebase or a specific file/module. Identifies bottlenecks, inefficient patterns, and optimization opportunities.
Modifies files
Hook triggers on file write and edit operations
Uses power tools
Uses Bash, Write, or Edit tools
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 claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
A stack-agnostic Claude Code plugin/template that turns the chat into a disciplined, spec-driven development pipeline: Spec Writer → Architect → Implementer, plus auto-invoked skills for bug fixing, frontend design, security/architecture/performance audits, and a meta skill to create new skills.
What you get:
spec-writer, architect, implementer) with clear roles and STOP protocols.AUTO_MODE=1 skips even that.WORKFLOW.md).The skill set is curated based on hard usage data — bugfix, frontend-design, and skill-creator made it in because they got used week after week, not because they sound nice.
Inside Claude Code:
/plugin marketplace add Valdemird/claude-structure
/plugin install claude-structure@claude-structure
The plugin auto-discovers agents/, skills/, and hooks/ (via hooks/hooks.json). Nothing else to wire up.
git clone https://github.com/Valdemird/claude-structure.git
cd claude-structure
# Copy plugin components into your project's .claude/
mkdir -p /path/to/your/project/.claude
cp -r agents skills hooks /path/to/your/project/.claude/
# Copy the project-level templates
cp templates/settings.json /path/to/your/project/.claude/settings.json
cp templates/CLAUDE.md.template /path/to/your/project/CLAUDE.md
$EDITOR /path/to/your/project/CLAUDE.md # describe your stack & commands
export TEST_CMD="pytest" # or "npx vitest run", "go test", etc.
export TEST_ALL_CMD="pytest"
export TYPECHECK_CMD="mypy"
export LINT_CMD="ruff check"
export FORMAT_CMD="black"
export AUTO_MODE=1 # optional: autonomous multi-phase runs
Open Claude Code in your project and the skills activate automatically based on the conversation.
claude-structure/ ← this is the plugin root
├── .claude-plugin/
│ ├── plugin.json ← plugin manifest
│ └── marketplace.json ← marketplace entry for /plugin marketplace add
├── agents/ ← plugin auto-discovers these
│ ├── architect.md
│ ├── implementer.md
│ └── spec-writer.md
├── skills/ ← plugin auto-discovers these
│ ├── spec-create/
│ ├── spec-audit/
│ ├── spec-review/
│ ├── spec-plan/
│ ├── spec-implement/
│ ├── spec-orchestrator/
│ ├── architecture/
│ ├── security-audit/
│ ├── performance/
│ ├── mobile-audit/
│ ├── bugfix/
│ ├── frontend-design/
│ └── skill-creator/
├── hooks/ ← plugin reads hooks.json + scripts
│ ├── hooks.json ← plugin hook config (uses ${CLAUDE_PLUGIN_ROOT})
│ ├── validate-spec-structure.sh
│ ├── validate-audit.sh
│ ├── validate-plan.sh
│ └── validate-implementation.sh
├── templates/ ← for copy-paste users only
│ ├── CLAUDE.md.template
│ └── settings.json
├── .github/workflows/validate.yml ← CI: shell, JSON, manifest, frontmatter
├── README.md
├── WORKFLOW.md
└── LICENSE
| Skill | Fires when you say… |
|---|---|
spec-create | "I want a spec for X", "create a spec for Y" |
spec-audit | "audit the spec for X" |
spec-review | "review the spec for X" |
spec-plan | "generate the plan for X", "plan the implementation" |
spec-implement | "implement phase N of X" (or end-to-end in AUTO_MODE=1) |
spec-orchestrator | "take this idea end-to-end as a spec" |
architecture | "do an architecture review", "audit the structure of this module" |
security-audit | "is this secure?", "check for XSS / SQLi / SSRF / auth issues" |
performance | "audit performance", "find bottlenecks" |
mobile-audit | "audit the mobile experience", "this is broken on iOS Safari" |
bugfix | "fix the bug where…", "X is broken / not working / regressed" |
frontend-design | "build me a landing page", "design this component" |
skill-creator | "create a new skill for…", "improve the X skill" |
npx claudepluginhub valdemird/claude-structure --plugin claude-structureHarness-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
Superpowers Plus core skills library for Claude Code: planning, execution routing, TDD, debugging, and collaboration workflows
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
Tools to maintain and improve CLAUDE.md files - audit quality, capture session learnings, and keep project memory current.
Comprehensive feature development workflow with specialized agents for codebase exploration, architecture design, and quality review
A growing collection of Claude-compatible academic workflow bundles. Covers scientific figures, manuscript writing and polishing, reviewer assessment, citation retrieval, data availability, paper reading, literature search, response letters, paper-to-PPTX conversion, and evidence-grounded Chinese invention patent drafting. Rules are organized as reusable skill folders with explicit workflows and quality checks.