claude-octopus
Multi-AI orchestration for research, code review, and shipping software
Tells the AI agent how to run Octopus workflows: which visual banners and provider indicators to display, how to route tasks across external AI providers, and which file-creation and cost-transparency rules to enforce during execution.
All Skills in This Repository (70)
Pure Emerald Level Indicatorsskill-finish-branch
Verify test results and finalize Git branches via merge, PR, keep, or discard.
skill-validate
Coordinate multi-AI debate to score code across four dimensions and extract issues.
skill-staged-review
Enforce two-stage review for spec compliance and code quality before merge.
skill-debate-integration
Automate multi-model AI debates with session-aware storage and quality gates.
sys-configure
Auto-configure Claude Octopus providers and preferences from the current environment.
skill-debate
Run structured three-way AI debates among Claude, Gemini, and Codex.
skill-tdd
Enforce a red-green-refactor test-driven workflow with mandatory verification phases.
skill-writing-plans
Convert vague specifications into executable implementation plans with explicit file paths, tests, and commit messages.
skill-status
Render a project status dashboard from .octo workflow state.
skill-audit
Identify and structure codebase issues through a repeatable auditing process.
flow-discover
Execute a multi-provider discovery workflow with Codex, Gemini, Perplexity, and Claude.
octopus-security-audit
Detect security vulnerabilities across codebases, deployment pipelines, and API surfaces.
Frequently Asked Questions
FAQPage SchemaHow to install claude-octopus?▼
Run `npx skills add nyldn/claude-octopus --all -g -y` in your terminal to install the full skill suite globally.
What does Claude Octopus actually do?▼
It coordinates multiple AI providers (Codex, Antigravity, Copilot, Ollama, and others) through a four-phase workflow — research, define, build, deliver — so no single model's blind spots reach your codebase.
Do I need paid API keys for every provider?▼
No. It works with Claude alone, and Codex, Antigravity, Copilot, and Ollama can use existing subscriptions or run locally at zero extra cost.
Can it review code and catch security issues?▼
Yes. It includes multi-LLM code review, OWASP security audits, adversarial red-team testing, and staged spec-vs-quality review pipelines with inline PR comments.
Does it work with tools other than Claude Code?▼
Yes. Skills ship in cross-platform formats for Codex, Cursor, OpenCode, and GitHub Copilot, plus an MCP server for any MCP-aware agent.
Related Repositories in Software Engineering
View All in Software Engineering→openclaw
Run a personal AI assistant across your devices and chat apps
superpowers
Gives coding agents a disciplined workflow from idea to merged code
react
AI agent skills for building, testing, and porting React core