Multi-model AI coding orchestrator where a coordinator routes Markdown-defined specialist tasks to Claude, Codex, Gemini, Kimi and Grok CLIs in isolated git worktrees.
rohitg00/pro-workflow
A Claude Code plugin with self-correcting memory, persistent FTS5 research wikis, and workflow skills, commands and hooks for AI coding agents.
- Stars
- 2,918
- Forks
- 310
- Open issues
- 32
- Last commit
- 20 days ago
What is it?
Pro Workflow is a plugin for Claude Code that stores corrections as rules in a SQLite database. Once a correction is saved, it is loaded at the start of later sessions so the same mistake is less likely to recur. Learnings are searchable with FTS5, and the store also holds persistent research wikis that can be queried from any session.
The plugin bundles 41 skills, 8 agents, 23 slash commands and 33 hook scripts across 22 lifecycle events. These cover end-of-session wrap-ups, multi-phase research-plan-implement-review flows, parallel git worktrees, permission denial analysis, MCP token overhead auditing and cost tracking. A budget-capped research loop can add pages to a wiki, and a multi-provider council can deliberate on a question and save the transcript as a wiki page.
Claude Code users install it through the plugin marketplace. The skills bundle can also be installed into other agents such as Cursor, Codex and Gemini CLI via the skills CLI.
Key features
- 01Saves corrections as rules in SQLite and loads them at session start.
- 02Provides FTS5-indexed research wikis that can be queried from any session.
- 03Includes 41 skills, 8 agents and 23 slash commands.
- 04Runs hook scripts across 22 Claude Code lifecycle events for quality gates and guardrails.
- 05Supports parallel git worktrees for running multiple sessions at once.
- 06Installs into other coding agents such as Cursor and Codex via skills add.
Use cases
- Stop repeating the same coding conventions and corrections across Claude Code sessions.
- Build a persistent, searchable research wiki on a technical topic over many sessions.
- Run several Claude Code sessions in parallel on separate git worktrees for large changes.
- Add commit, secret-scanning and destructive-command guardrails to an AI coding workflow.
Topics
FAQ
- What is pro-workflow?
- A Claude Code plugin with self-correcting memory, persistent FTS5 research wikis, and workflow skills, commands and hooks for AI coding agents. Pro Workflow is a plugin for Claude Code that stores corrections as rules in a SQLite database. Once a correction is saved, it is loaded at the start of later sessions so the same mistake is less likely to recur. Learnings are searchable with FTS5, and the store also holds persistent research wikis that can be queried from any session.
- Is pro-workflow free to use? What license does it have?
- pro-workflow has no declared license. Its code is public, but without a license you have no explicit permission to reuse it; ask the author before relying on it.
- Is pro-workflow still maintained?
- Actively maintained. The last commit to pro-workflow was on Sep 29, 2026.
- What can I use pro-workflow for?
- Stop repeating the same coding conventions and corrections across Claude Code sessions. Build a persistent, searchable research wiki on a technical topic over many sessions. Run several Claude Code sessions in parallel on separate git worktrees for large changes. Add commit, secret-scanning and destructive-command guardrails to an AI coding workflow.
- Where can I get pro-workflow?
- The source code and installation instructions are on GitHub: https://github.com/rohitg00/pro-workflow
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
A CLI and workflow engine that runs coding agents like Claude Code, Codex and Copilot through configurable review, repair and test gates before work is accepted.
Desktop, mobile, web and CLI app for running and orchestrating multiple coding agents such as Claude Code, Codex and OpenCode in parallel worktrees.
Two free online books explaining how agent harnesses for code-writing models, such as Claude Code and Codex, are designed and controlled.
A prompt-based skill that makes AI coding agents write only the code a task needs, favoring reuse and minimal solutions over large implementations.
A distributed, Dolt-backed graph issue tracker that gives coding agents persistent, dependency-aware task memory via the bd CLI.