Imported from ecasilla/super-opencode (
AGENTS.md). Install upstream withnpx skills add ecasilla/super-opencode. Copyright stays with the author.
SuperClaude → OpenCode Framework
This configuration brings SuperClaude's specialized agent system to OpenCode with 20 agents and 28 commands.
Agent Architecture
Agents are organized by domain and follow the principle of least privilege for tool access.
Architecture & Design (read-only analysis)
- backend-architect — Backend systems, APIs, databases, security
- frontend-architect — UI/UX, accessibility (WCAG), Core Web Vitals
- system-architect — Scalable architecture, dependency management, patterns
Security & Performance (analysis + investigation)
- security-engineer — Vulnerability assessment, OWASP, threat modeling, compliance
- performance-engineer — Measurement-driven optimization, benchmarking, profiling
Research & Analysis (read + web access)
- deep-research-agent — Comprehensive research with adaptive strategies
- deep-research — External knowledge gathering with credibility assessment
- root-cause-analyst — Evidence-based problem investigation, hypothesis testing
- requirements-analyst — Requirements discovery, PRD creation, scope definition
Development (full code access)
- python-expert — Production-ready Python with SOLID, TDD, security-first
- refactoring-expert — Code quality improvement, technical debt reduction
- devops-architect — CI/CD, IaC, observability, container orchestration
Quality & Review
- quality-engineer — Testing strategies, edge cases, automation, coverage
- self-review — Post-implementation validation, checklist-style reports
Education & Communication
- learning-guide — Progressive programming education with examples
- socratic-mentor — Socratic method for discovery-based learning
- technical-writer — Audience-targeted documentation
Strategy & Project Management
- business-panel-experts — 9-expert strategy panel (Christensen, Porter, Drucker, etc.)
- pm-agent — PDCA workflow, memory integration, knowledge management
- repo-index — Repository indexing for token-efficient context
Command Quick Reference
| Command | Description | Default Agent |
|---|---|---|
/analyze |
Code quality/security/performance analysis | current |
/brainstorm |
Requirements discovery via Socratic dialogue | requirements-analyst |
/build |
Project compilation and packaging | devops-architect |
/business-panel |
Multi-expert business strategy analysis | business-panel-experts |
/cleanup |
Dead code removal, import optimization | refactoring-expert |
/design |
System/component architecture design | system-architect |
/document |
Documentation generation | technical-writer |
/estimate |
Time/effort/complexity estimation | current |
/explain |
Code and concept explanation | learning-guide |
/git |
Intelligent Git operations | current |
/help |
Command help and guidance | current |
/implement |
Code implementation from specs | python-expert |
/improve |
Code quality improvement | refactoring-expert |
/index-repo |
Repository indexing | repo-index |
/load |
Restore session context | pm-agent |
/pm |
Project management (PDCA) | pm-agent |
/recommend |
Technology recommendations | system-architect |
/reflect |
Post-implementation reflection | self-review |
/research |
Deep research with citations | deep-research-agent |
/save |
Save session state | pm-agent |
/select-tool |
Tool recommendations | current |
/spawn |
Spawn specialized subagent | current |
/spec-panel |
Technical specification review | current |
/task |
Task management and tracking | pm-agent |
/test |
Testing strategy and execution | quality-engineer |
/troubleshoot |
Problem investigation | root-cause-analyst |
/workflow |
Multi-step workflow orchestration | pm-agent |
/agent |
Agent management | current |
Workflow Patterns
New Feature: /brainstorm → /design → /implement → /test → /reflect
Bug Fix: /troubleshoot → /implement → /test → /reflect
Code Quality: /analyze → /improve → /cleanup → /test
Research: /research → /recommend → /design
Project Kickoff: /index-repo → /brainstorm → /estimate → /workflow