Workflows

A curated collection of development, data science, writing, workshop, legal, and research workflows for Claude Code.

Quick Start

# 1. Install the plugin
/plugin marketplace add edwinhu/workflows

# 2. (Optional) Install CLI dependencies for knowledge management, email, and calendar
bash ~/.claude/plugins/cache/edwinhu-plugins/workflows/*/bin/install-deps.sh

The core workflows (/dev, /ds, /writing, /workshop, document formats) work without any CLI dependencies. The install script is only needed for skills that use external tools.

This downloads pre-built binaries for your platform (macOS arm64, Linux x64, Windows x64) from GitHub Releases:

Tool	Purpose	Used by
nlm	NotebookLM CLI	librarian agent, /nlm
readwise-custom	Readwise RAG/chat/upload	librarian agent, /readwise-chat
scholar	Google Scholar search	librarian agent, /google-scholar
consensus	Academic paper search	librarian agent, /consensus
morgen	Calendar & tasks	Direct Bash, or session in ~/areas/assistant/
superhuman	Email	email-handling skill (via Bash)

Requires gh (GitHub CLI). Tools already on your $PATH are skipped.

---

User Commands (15)

These are the skills you invoke directly with /name:

Core Workflows

Four primary workflows, each with a fresh-start entry and a midpoint re-entry:

Start Fresh	Midpoint Re-entry	Domain
/dev	/dev-debug	7-phase feature development with TDD enforcement
/ds	/ds-fix	6-phase data science with output-first verification and DQ validation
/writing	/writing-revise	6-phase writing with claim validation and deviation rules
/workshop	/workshop-revise	4-phase workshop presentations from research papers

Document Formats

Skill	Purpose
/docx	Word document creation, editing, tracked changes
/pdf	PDF extraction, creation, form filling
/pptx	Presentation creation and editing
/xlsx	Spreadsheet creation and analysis

Office format skills sourced from anthropics/skills via git submodule.

Meta

Skill	Purpose
/skill-creator	Skill creation with superpowers enforcement patterns
/workflow-creator	Design structured multi-phase LLM workflows
/headline-card	LWT-style headline cards for Typst presentations

---

Auto-Invoked Skills (60+)

These skills have user-invocable: false — Claude loads them automatically when relevant. You don't call them directly.

Legal & Citation

bluebook, bluebook-audit, docx-footnotes, source-verify

Data Access

wrds, lseg-data, gemini-batch

Knowledge Management

nlm, google-scholar, readwise, readwise-chat, readwise-search, readwise-docs, readwise-prune

Research

consensus, research

Notebook Tools

marimo, jupytext, notebook-debug

Utilities

look-at, visual-verify, visual-mockup, data-context, continuous-learning, pattern-capture, ai-anti-patterns, dev-tools, ds-tools, dev-worktree, obsidian-organize, pptx-render, audit-fix-loop, plugin-creator

Internal Workflow Phases

Dev: dev-clarify, dev-explore, dev-design, dev-delegate, dev-implement, dev-tdd, dev-review, dev-verify, dev-handoff, dev-spec-reviewer, dev-plan-reviewer, dev-test, dev-test-* (cross-turn iteration uses Claude Code's built-in /goal)

DS: ds-plan, ds-delegate, ds-implement, ds-review, ds-validate, ds-verify, ds-handoff, ds-spec-reviewer, ds-plan-reviewer

Writing: writing-setup, writing-outline, writing-outline-reviewer, writing-precis-reviewer, writing-draft, writing-econ, writing-general, writing-legal, writing-review, writing-validate, writing-handoff

Workshop: (no internal phase skills — workshop and workshop-revise are both user-facing entry points)

---

Agents (20)

Specialized subagents auto-discovered by Claude Code from agents/:

Agent	Role
planner	Implementation planning for complex features
architect	System design and technical decisions
tdd-guide	TDD workflow enforcement
dev-implementer	Feature implementation with automatic linting
dev-debugger	Hypothesis-driven debugging with serial iteration
dev-verifier	Goal-backward verification (4-level: exists, substantive, wired, functional)
dev-plan-checker	Plan review before implementation
test-gap-auditor	Requirement-to-test coverage mapping and gap filling
ds-analyst	Data analysis with output-first verification
ds-engineer	Data engineering pipelines, ETL, and transformations
code-reviewer	Code quality, security, and maintainability review
security-reviewer	Security vulnerability detection
build-error-resolver	Fix build/type errors with minimal diffs
e2e-runner	Playwright E2E testing
refactor-cleaner	Dead code cleanup and consolidation
doc-updater	Documentation sync and codemap updates
data-explorer	EDA and data profiling
librarian	Knowledge management orchestration (NLM, Readwise, Scholar)

---

Hooks (11)

Hooks auto-run at specific lifecycle events:

Script	Event	Trigger	Purpose
session-start.py	SessionStart	startup/resume/clear/compact	Inject using-skills meta-skill
session-end.py	Stop	*	Update LEARNINGS.md timestamp
pre-compact.py	PreCompact	*	Preserve state before compaction
suggest-compact.py	PreToolUse	Edit/Write	Suggest compaction when context is large
image-read-guard.py	PreToolUse	Read	Redirect to look-at for media files
lint-check.py	PostToolUse	Edit/Write	Lint after file changes (ESLint, ruff, lintr)
writing-suggest-verify.py	PostToolUse	Edit/Write	Suggest visual verification for writing
atomic-constraint-guard.py	PostToolUse	Edit/Write	Validate atomic constraint file structure
pr-url-logger.py	PostToolUse	Bash	Log PR URLs and GitHub Actions status
overflow-check.py	PostToolUse	Bash	Detect Typst content overflow after compilation
pattern-scan.py	SessionEnd	clear/logout	Scan session for reusable patterns

---

Session Continuity

Workflows use a .planning/ state folder (SPEC.md, PLAN.md, LEARNINGS.md, HANDOFF.md, VALIDATION.md) for cross-session persistence. The /continuous-learning skill extracts reusable patterns, and hooks automatically save session state.

For cross-session task persistence, set CLAUDE_CODE_TASK_LIST_ID in .envrc:

export CLAUDE_CODE_TASK_LIST_ID="my-project"

---

Repository Structure

workflows/
├── .claude-plugin/             # Plugin manifest
│   ├── plugin.json             # Version and metadata
│   └── marketplace.json        # Marketplace listing
├── agents/                     # Specialized subagents (19)
├── skills/                     # All skills (15 user-facing + 60+ internal)
│   ├── dev/, ds/, writing*/, workshop*/  # Core workflow entry points
│   ├── docx, pdf, pptx, xlsx  # Document formats (symlinks)
│   └── ...                     # Internal phases and auto-invoked skills
├── hooks/                      # Hook scripts (11)
│   ├── hooks.json              # Hook configuration
│   └── *.py                    # Hook implementations
├── references/                 # Shared constraint and reference docs
├── external/
│   └── anthropic-skills/       # Git submodule for document skills
└── PHILOSOPHY.md               # Workflow design philosophy

Key Points:

• skills/ contains both user-facing and internal phase skills (auto-discovered; internal skills use user-invocable: false)
• agents/ contains specialized subagents (19, auto-discovered by Claude Code)
• hooks/ contains hook entry points called directly by hooks.json
• references/ contains shared constraint and enforcement docs

Updating External Skills

The office format skills come from Anthropic's official skills repo. To update:

git submodule update --remote external/anthropic-skills

Acknowledgments

This project was heavily inspired by obra/superpowers, particularly:

• The SessionStart hook pattern for injecting meta-skills
• The "using-skills" approach that teaches HOW to use skills rather than listing WHAT skills exist
• The philosophy that skills should be invoked on-demand, not dumped into every session

Office format skills (docx, pdf, pptx, xlsx) are from anthropics/skills.

License

MIT

Author

Edwin Hu