opencode-agents
Multi-agent coding workflows with orchestration, review, and debugging
All Skills in This Repository (40)
Pure Emerald Level Indicatorssenior-security
Coordinates security workflows for architecture design, threat modeling, auditing, and cryptography across SDLC phases.
using-git-worktrees
Create isolated git worktrees with priority-based directory selection and safety checks.
cynefin
Classify problems into Cynefin domains and recommend management responses.
git-commit
Enforce Conventional Commits format with validated types and subject limits.
moscow
Classify backlog items into Must, Should, Could, and Won't categories.
5whys
Guide iterative Why questioning to identify root causes and corrective actions.
test-driven-development
Enforce test-first development with failing tests before minimal code.
agent-tuning
Configure and optimize OpenCode/Claude coding agents with prompts, permissions, and workflows.
systematic-debugging
Guide software debugging through four mandatory investigation phases before fixes.
cto-advisor
Provides strategic technology leadership guidance for engineering organizations.
rice
Score initiatives with RICE inputs and output a ranked list.
redteam
Analyze plans and systems to identify vulnerabilities and failure modes.
Frequently Asked Questions
FAQPage SchemaHow to install opencode-agents?βΌ
Run `npx skills add wildwasser/opencode-agents --all -g -y` in your terminal to install all skills globally.
What is the multi-agent workflow in opencode-agents?βΌ
An orchestrator agent (Oscar) delegates research to Scout, implementation to Ivan, and adversarial review to Jester, keeping each agent's context lean and code quality high.
Does opencode-agents work with Claude Code?βΌ
Yes. All skills follow the universal SKILL.md standard and work in Claude Code, opencode, Cursor, and other compatible coding environments.
How does it prevent bad code from being committed?βΌ
Built-in skills enforce test-driven development, systematic root-cause debugging, spec compliance review, and verification with real command output before any completion claim.
Can I use it for planning and prioritization?βΌ
Yes. It includes decision frameworks like MoSCoW, RICE, SWOT, Cynefin, OODA, pre-mortem, and red team analysis for planning and risk review.
Related Repositories in Software Engineering
View All in Software Engineeringβopenclaw
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