Continuous-Claude-v3
Continuous-Claude-v3 is an open-source Python project in the claude space with 3.8k GitHub stars. The repository has seen commits within the last year.
MITPermissive — free to use in commercial and proprietary software, with attribution.View license →
Production readiness
4/5- Actively maintainedCommits in the last 6 months
- No known vulnerabilitiesNot yet scanned
- Clear, usable licenseMIT (permissive)
- Proven adoptionWidely used
- Has documentationDocumentation indexed
pip install continuous-claude-v3OPC-Dev Documentation
Extended Claude Code workspace with workflow orchestration, specialist agents, and development automation.
Quick Start
/help # Interactive discovery - start here
Documentation Structure
docs/
├── README.md # This file
├── ARCHITECTURE.md # System architecture and data flow
├── workflows/ # Workflow orchestration
│ └── README.md # /fix, /build, /tdd, /prove workflows
├── agents/ # Specialist agents
│ └── README.md # 31 agents by category
├── hooks/ # Hook system
│ └── README.md # Lifecycle hooks and custom hooks
├── tools/ # CLI tools
│ └── README.md # tldr, memory, MCP integrations
└── skills/ # Skill documentation
└── help.md # /help discovery skill
Quick Links
Workflows - Multi-agent pipelines (/fix, /build, /tdd)
Agents - 31 specialist agents
Hooks - Automatic behaviors and custom hooks
Tools - TLDR CLI, /prove, memory system, MCP
Architecture - System design and data flow
What's Included
Workflows
Multi-agent pipelines for complex tasks:
/fix- Bug investigation → diagnosis → implementation/build- Feature planning → implementation → testing/tdd- Test-driven development cycle/prove- Formal math proofs with Lean 4
Agents (31 specialists)
Exploration: scout, oracle, pathfinder
Planning: architect, plan-agent, phoenix
Implementation: kraken (TDD), spark (quick fixes)
Review: arbiter, critic, judge
Investigation: sleuth, debug-agent, profiler
Tools
tldr- Token-efficient code analysis (95% savings)Memory system - Store/recall learnings across sessions
Premortem - Risk analysis before implementation
Hooks (25+ active)
Lifecycle extensions for validation, coordination, and automation.
Getting Started
Run
/helpto discover capabilities interactivelyDescribe your goal - Claude routes to the right workflow
Or invoke specific workflows:
/fix,/build,/prove
Requirements
Claude Code CLI
Node.js (for hooks)
Python 3.10+ with uv (for scripts)
Optional: LM Studio with Godel-Prover (for /prove)