CoCo Super Intelligence — Orchestrates AI agents into an "engineering department" with personas, skills, and persistent state for better dev decisions.
Analyzed by Sai Pavan Gopularam · AI · Developer Tools · View on GitHub
- Stars: 660
- Forks: 12
- Commits last 30 days: 100
- Health: Active (100 commits this month)
- Language: Python
- License: Other
What It Is
Imagine you're trying to make a tough technical decision, and instead of just one AI's opinion, you get a panel of 495 "expert" personas, each debating the issue from different angles, citing sources, and even providing dissent. CoCo Super Intelligence is that orchestration layer, turning your existing AI coding tools (like Claude Code or Cursor) into a sophisticated advisory board and task execution engine. It's like having a virtual R&D department at your fingertips.
This system solves the problem of getting generic, uncontextualized AI advice. Instead of a single, often bland, response, CoCo provides a structured debate, named dissent, and persistent state for decisions and tasks. It helps developers make more informed decisions and then execute on them with a vast catalog of skills, all while keeping your data local.
License Verdict
Mixed License — Core MIT, Super Intelligence Proprietary — Commercial Use for Core • Super Intelligence is Own-Use Only
The core components of CoCo are released under the permissive MIT License, allowing for free use, modification, and distribution, including commercial use. However, the "Super Intelligence" features, which include the 495 personas and the `/SI-*` commands, are explicitly proprietary and designated for "own-use." This means you cannot offer the Super Intelligence features as a commercial service or product to external customers.
How to Use It
Get started by installing CoCo via `npx`, which clones the repository and runs an install script. It integrates with your existing AI coding environments like Claude Code or Cursor, allowing you to immediately use its decision-making and skill execution capabilities.
Prerequisites:
- git
- bash
- Node.js 14+
- python3
Estimated setup time: 3 minutes.
npx cocosuperintelligence
# Then, in Claude Code:
/SI-Decide "Should a 10-person startup move its public REST API to GraphQL this quarter?"What I'd Build With This
Internal AI Workflow Automation Toolkit (micro-saas)
Develop a specialized internal toolkit for small development agencies or consulting firms. This toolkit would leverage CoCo's MIT-licensed core skills for code generation, refactoring, and project management, integrated with existing developer tools. The "Super Intelligence" features would be used internally by the agency's developers to enhance their own productivity and decision-making on client projects, not directly sold as a service.
Effort: 1 Week Build Time · Target: Small Dev Agencies · Pricing: $199/month per team license
AI-Powered Technical Consulting Platform (saas)
Build a consulting platform where engineering teams can submit complex technical dilemmas. Your team of expert consultants would then use CoCo's "Super Intelligence" features internally to generate comprehensive, debated recommendations and implementation plans. The platform would deliver these detailed reports and execution strategies to clients, effectively productizing the *output* of CoCo's SI, rather than the SI itself as a direct service.
Effort: 3 Months Build Time · Target: Mid-Market Engineering Teams · Pricing: $500 - $5,000 per project
Internal AI Decision & Execution Engine for Large Orgs (enterprise)
Offer a tailored deployment and integration service for large enterprises looking to enhance their internal engineering and product development processes. You would deploy CoCo's full "Super Intelligence" system on-premise or within the enterprise's private cloud, integrating it deeply with their existing codebases, documentation, and development workflows. This allows the enterprise to leverage the 495 personas and 342 commands for internal decision-making and task automation, fully compliant with the "own-use" license.
Effort: 6 Months + Customization · Target: Fortune 500 Tech Departments · Pricing: $50,000 - $250,000 per deployment
Sai Pavan Gopularam's Take
This is fascinating. The idea of an AI orchestrating a debate among 'expert' personas to reach a decision is a novel approach to prompt engineering. The 'own-use' license for the Super Intelligence part means you can't build a direct SaaS around the personas, but I could see a consulting business charging $10,000 per engagement to leverage this internally for clients.
Watch Out For
- External GSD Toolkit Dependency: Many skills within the GSD (Get Shit Done) bundle require an upstream GSD toolkit at `~/.claude/get-shit-done`, which is not bundled with the current installer. This means the full functionality of GSD skills might not be immediately available.
- Persona Disclaimer: The 495 personas are modeled on public writing and are not endorsements or actual views of the named experts. While cited, the system confirms a stance is *cited*, not that a quotation is authentic.
I break down trending repos like CoCo Super Intelligence every week — join the newsletter.