pro-workflow

Pro Workflow

Battle-tested AI coding workflows from power users. Self-correcting memory, parallel worktrees, wrap-up rituals, and the 80/20 AI coding ratio. Works with Claude Code and Cursor.

The Core Idea

"80% of my code is written by AI, 20% is spent reviewing and correcting it." — Karpathy

This skill optimizes for that ratio. Every pattern reduces correction cycles.

Patterns

Pattern	What It Does
Self-Correction Loop	Claude learns from your corrections automatically
Parallel Worktrees	Zero dead time - native claude -w worktrees
Wrap-Up Ritual	End sessions with intention, capture learnings
Split Memory	Modular CLAUDE.md for complex projects
80/20 Review	Batch reviews at checkpoints
Model Selection	Opus 4.6 adaptive thinking, Sonnet 4.6 (1M context)
Context Discipline	Manage your 200k token budget
Learning Log	Auto-document insights

Installation

Cursor (Recommended)

/add-plugin pro-workflow

The plugin includes 9 skills, 3 agents, and 6 rules that load automatically.

Claude Code — One-Click Plugin Install

# Add marketplace
/plugin marketplace add rohitg00/pro-workflow

# Install plugin
/plugin install pro-workflow@pro-workflow

Or via CLI:

claude plugin marketplace add rohitg00/pro-workflow
claude plugin install pro-workflow@pro-workflow

Claude Code — Build with SQLite Support

After installation, build the TypeScript for persistent storage:

cd ~/.claude/plugins/*/pro-workflow  # Navigate to plugin directory
npm install && npm run build

This creates the SQLite database at ~/.pro-workflow/data.db.

Or load directly

claude --plugin-dir /path/to/pro-workflow

Manual Setup

git clone https://github.com/rohitg00/pro-workflow.git /tmp/pw
cp -r /tmp/pw/templates/split-claude-md/* ./.claude/
cp -r /tmp/pw/commands/* ~/.claude/commands/
cp -r /tmp/pw/hooks/* ~/.claude/

Minimal (Just add to CLAUDE.md)

## Pro Workflow

### Self-Correction
When corrected, propose rule → add to LEARNED after approval.

### Planning
Multi-file: plan first, wait for "proceed".

### Quality
After edits: lint, typecheck, test.

### LEARNED

Skills (Cursor)

Skill	Description
pro-workflow	Core 8 workflow patterns for AI-assisted coding
smart-commit	Quality gates, staged review, and conventional commits
wrap-up	End-of-session ritual with change audit and learning capture
learn-rule	Capture corrections as persistent learning rules
parallel-worktrees	Set up git worktrees for zero dead time
replay-learnings	Surface past learnings relevant to the current task
session-handoff	Generate handoff documents for session continuity
insights	Session analytics, correction trends, and productivity metrics
deslop	Remove AI-generated code slop and clean up style

Rules (Cursor)

Rule	Applies To	Description
quality-gates	Always	Lint, typecheck, and test before commits
atomic-commits	Always	Conventional format, feature branches, specific staging
context-discipline	Always	Read before edit, plan before multi-file changes
self-correction	Always	Capture mistakes as compounding learnings
no-debug-statements	*.{ts,tsx,js,jsx,py,go,rs}	Remove console.log, debugger, print before committing
communication-style	Always	Concise, action-oriented, no over-engineering

Commands (Claude Code)

After plugin install, commands are namespaced:

Command	Purpose
/pro-workflow:wrap-up	End-of-session checklist
/pro-workflow:learn-rule	Extract correction to memory (file-based)
/pro-workflow:parallel	Worktree setup guide
/pro-workflow:learn	Claude Code best practices & save learnings
/pro-workflow:search	Search learnings by keyword
/pro-workflow:list	List all stored learnings
/pro-workflow:commit	Smart commit with quality gates and code review
/pro-workflow:insights	Session analytics, learning patterns, and correction heatmap
/pro-workflow:replay	Surface past learnings for current task
/pro-workflow:handoff	Generate session handoff document for next session

Database Features

Persistent Learnings

Learnings are stored in SQLite with FTS5 full-text search:

~/.pro-workflow/
└── data.db    # SQLite database with learnings and sessions

Search Examples

/search testing           # Find all testing-related learnings
/search "file paths"      # Exact phrase search
/search git commit        # Multiple terms

Learning Categories

• Navigation (file paths, finding code)
• Editing (code changes, patterns)
• Testing (test approaches)
• Git (commits, branches)
• Quality (lint, types, style)
• Context (when to clarify)
• Architecture (design decisions)
• Performance (optimization)
• Claude-Code (sessions, modes, CLAUDE.md, skills, subagents, hooks, MCP)
• Prompting (scope, constraints, acceptance criteria)

Hooks

Automated enforcement of workflow patterns.

Hook	When	What
PreToolUse	Before edits	Track edit count, quality gate reminders
PreToolUse	Before git commit/push	Remind about quality gates, wrap-up
PostToolUse	After code edits	Check for console.log, TODOs, secrets
PostToolUse	After tests	Suggest [LEARN] from failures
UserPromptSubmit	Each prompt	Drift detection — warns when straying from original intent
SessionStart	New session	Load learnings from database
Stop	Each response	Context-aware reminders using last_assistant_message
SessionEnd	Session close	Save session stats to database
ConfigChange	Settings modified	Detect when quality gates or hooks are changed mid-session

Install Hooks

# Full setup with hooks
git clone https://github.com/rohitg00/pro-workflow.git /tmp/pw
cp -r /tmp/pw/hooks/* ~/.claude/
cp -r /tmp/pw/scripts ~/.claude/scripts/pro-workflow/
cp -r /tmp/pw/commands/* ~/.claude/commands/

Contexts & Agents

Context	When	Behavior
dev	Building	Code first, iterate
review	PRs	Read-only, security focus
research	Exploring	Summarize, plan

Agent	Purpose
planner	Break down complex tasks
reviewer	Code review, security audit
scout	Background confidence-gated exploration with worktree isolation

Agent Teams (Experimental)

Coordinate multiple Claude Code sessions working together:

# Enable in settings.json
{ "env": { "CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS": "1" } }

• Lead session coordinates, teammates work independently
• Teammates message each other directly
• Shared task list with dependency management
• Display modes: in-process (Shift+Down to navigate, wraps around) or split panes (tmux/iTerm2)
• Delegate mode (Shift+Tab): lead orchestrates only
• Docs: https://code.claude.com/docs/agent-teams

Structure

pro-workflow/
├── .claude-plugin/              # Claude Code plugin
│   ├── plugin.json
│   ├── marketplace.json
│   ├── settings.json            # Default permissions for quality gates
│   └── README.md
├── .cursor-plugin/              # Cursor plugin
│   └── plugin.json
├── skills/                      # Shared skills (Cursor + Claude Code)
│   ├── pro-workflow/SKILL.md
│   ├── smart-commit/SKILL.md
│   ├── wrap-up/SKILL.md
│   ├── learn-rule/SKILL.md
│   ├── parallel-worktrees/SKILL.md
│   ├── replay-learnings/SKILL.md
│   ├── session-handoff/SKILL.md
│   ├── insights/SKILL.md
│   └── deslop/SKILL.md
├── agents/                      # Shared agents (Cursor + Claude Code)
│   ├── planner.md
│   ├── reviewer.md
│   └── scout.md
├── rules/                       # Rules
│   ├── core-rules.md            # Claude Code rules
│   ├── quality-gates.mdc        # Cursor rules
│   ├── atomic-commits.mdc
│   ├── context-discipline.mdc
│   ├── self-correction.mdc
│   ├── no-debug-statements.mdc
│   └── communication-style.mdc
├── commands/                    # Claude Code commands
│   ├── wrap-up.md
│   ├── learn-rule.md
│   ├── commit.md
│   ├── insights.md
│   ├── replay.md
│   ├── handoff.md
│   └── ...
├── hooks/                       # Claude Code hooks
│   └── hooks.json
├── scripts/                     # Hook scripts (includes config-watcher.js)
├── contexts/                    # Context modes
│   ├── dev.md
│   ├── review.md
│   └── research.md
├── src/                         # TypeScript source (SQLite)
│   ├── db/
│   └── search/
├── assets/
│   └── logo.svg                 # Plugin logo
├── package.json
├── tsconfig.json
└── README.md

Learn Claude Code

Pro-workflow teaches Claude Code best practices directly, with links to official documentation for deep dives.

Official Docs: https://code.claude.com/docs/

Topics covered: sessions, context management, modes, CLAUDE.md, prompting, writing rules, skills, subagents, hooks, MCP, security, and IDE integration.

/pro-workflow:learn                 # Best practices guide & save learnings
/pro-workflow:learn-rule            # Capture corrections to memory
/pro-workflow:search claude-code    # Find past Claude Code learnings

SkillKit - Universal AI Skills

One skill. 32+ AI coding agents. Install pro-workflow across Claude Code, Cursor, Codex, Gemini CLI, and more with SkillKit.

# Install this skill
npx skillkit install pro-workflow

# Translate to any agent format
npx skillkit translate pro-workflow --agent cursor

# Get AI-powered skill recommendations
npx skillkit primer

Why SkillKit?

• Install once, use everywhere
• 15,000+ skills in the marketplace
• Works with Claude Code, Cursor, Codex, Gemini CLI, Windsurf, and 27+ more agents

Explore the marketplace at agenstskills.com

Philosophy

• Compound improvements - Small corrections → big gains over time
• Trust but verify - Let AI work, review at checkpoints
• Zero dead time - Parallel sessions keep momentum
• Memory is precious - Both yours and Claude's

Support

If you find this useful:

• Star this repo to help others discover it
• Check out SkillKit for more AI coding skills
• Report issues or suggest improvements

Distilled from Claude Code power users and real production use.