Eoc Opencode

Note: This is a forked version of everything-claude-code specifically adapted for OpenCode.

Everything OpenCode (EOC) provides a comprehensive plugin for OpenCode with agents, commands, skills, hooks, and rules for AI-assisted software development.

Features

Feature	Count	Description
Agents	16	Specialized subagents for delegation
Commands	40+	Slash commands for quick execution
Hooks	11+	Event-based automations
Custom Tools	3	Native tools for testing and security
Instructions	1	Consolidated rules and guidelines

Installation

Quick Install

# Global install (to ~/.config/opencode/)
bun x eoc-opencode@latest

# Local install (to project .opencode/ directory)
bun x eoc-opencode@latest --local

# Upgrade existing
bun x eoc-opencode@latest --upgrade

# Uninstall
bun x eoc-opencode@latest --uninstall

---

How to Use

Using Commands

Commands are invoked with / prefix. They delegate to specialized agents.

# Plan a feature
/plan

# Run TDD workflow
/tdd

# Review code
/code-review

Using Agents Directly

You can invoke agents directly via the /ask command or by mentioning them:

# Use planner agent directly
@planner Design a user authentication system

# Use security reviewer
@security-reviewer Audit the auth module

Using Skills

Skills provide domain-specific knowledge. Load them when needed:

# Load backend patterns skill
@skill backend-patterns

# Load security review skill  
@skill security-review

---

Available Commands

Planning & Development

Command	Description	Usage Example
/plan	Create implementation plan	/plan Build a user authentication system
/tdd	Test-driven development	/tdd Add login validation
/code-review	Code quality review	/code-review or /code-review --security
/build-fix	Fix build errors	/build-fix
/refactor-clean	Remove dead code	/refactor-clean

Testing & Quality

Command	Description	Usage Example
/e2e	Generate E2E tests	/e2e User login flow
/test-coverage	Analyze coverage	/test-coverage
/verify	Verification loop	/verify

Security

Command	Description	Usage Example
/security-scan	Security audit	/security-scan

Language-Specific

Command	Description	Usage Example
/go-review	Go code review	/go-review
/go-build	Fix Go build errors	/go-build
/go-test	Run Go tests	/go-test
/python-review	Python code review	/python-review

Multi-Model Operations

Command	Description	Usage Example
/multi-plan	Multi-model planning	/multi-plan Feature request
/multi-execute	Multi-model execution	/multi-execute Implement auth
/multi-frontend	Frontend multi-model	/multi-frontend React component
/multi-backend	Backend multi-model	/multi-backend API endpoint
/multi-workflow	Workflow multi-model	/multi-workflow Full feature

Documentation & Refactoring

Command	Description	Usage Example
/update-docs	Update documentation	/update-docs
/refactor-clean	Clean dead code	/refactor-clean

Operations & DevOps

Command	Description	Usage Example
/pm2	PM2 service management	/pm2 status
/setup-pm	Setup PM2	/setup-pm
/sessions	Session history	/sessions
/harness-audit	Audit harness	/harness-audit
/quality-gate	Quality checks	/quality-gate

Loops & Automation

Command	Description	Usage Example
/loop-start	Start agentic loop	/loop-start Build feature
/loop-status	Check loop status	/loop-status
/evolve	Auto-evolve code	/evolve

Learning & Evaluation

Command	Description	Usage Example
/learn	Learn from session	/learn
/learn-eval	Evaluate learning	/learn-eval
/eval	Evaluate performance	/eval
/checkpoint	Save checkpoint	/checkpoint

Model Management

Command	Description	Usage Example
/model-route	Route to models	/model-route

Project Management

Command	Description	Usage Example
/projects	List projects	/projects
/promote	Promote changes	/promote

Instincts

Command	Description	Usage Example
/instinct-status	Check instincts	/instinct-status
/instinct-export	Export instincts	/instinct-export
/instinct-import	Import instincts	/instinct-import

---

Available Agents

Planning & Architecture

Agent	Description	How to Use
planner	Creates detailed implementation plans	@planner Design a REST API
architect	System design and architecture	@architect Scalable microservices

Development

Agent	Description	How to Use
tdd-guide	Enforces test-first development	@tdd-guide Implement user auth
build-error-resolver	Fixes build/compilation errors	@build-error-resolver

Review & Quality

Agent	Description	How to Use
code-reviewer	Code quality and security review	@code-reviewer
security-reviewer	Vulnerability analysis	@security-reviewer
e2e-runner	Playwright E2E tests	@e2e-runner Login flow
refactor-cleaner	Dead code removal	@refactor-cleaner
doc-updater	Documentation sync	@doc-updater

Language-Specific

Agent	Description	How to Use
go-reviewer	Go code review	@go-reviewer
go-build-resolver	Go build errors	@go-build-resolver
python-reviewer	Python code review	@python-reviewer
database-reviewer	Supabase/PostgreSQL	@database-reviewer

Operations

Agent	Description	How to Use
chief-of-staff	Communication triage	@chief-of-staff
loop-operator	Autonomous loops	@loop-operator
harness-optimizer	Config optimization	@harness-optimizer

---

Available Skills

Skills are loaded on-demand for domain-specific guidance.

core Skills

Skill	Description	When to Use
coding-standards	TS/JS/React best practices	General development
backend-patterns	Node.js/Express/Next.js APIs	Backend development
frontend-patterns	React/Next.js patterns	Frontend development
security-review	Security checklists	Auth, payments, APIs
tdd-workflow	TDD methodology	Test-driven development
continuous-learning	Pattern extraction	Learning from sessions

Loading a Skill

# The skill tool loads automatically when you mention the skill
# e.g., "Use the security-review skill" loads it automatically

# Or explicitly:
@skill security-review

---

Custom Tools

run-tests

Run test suite with optional coverage:

await runTests({ coverage: true, file: "src/auth.test.ts" })

check-coverage

Verify 80%+ test coverage:

await checkCoverage({ threshold: 80, reporter: "text" })

security-audit

Run security scan:

await securityAudit({ fix: true, deep: false })

---

Hook Events

The plugin supports OpenCode events:

Event	Description
session.created	Session initialization
session.idle	Idle detection
session.deleted	Cleanup
tool.execute.before	Pre-tool execution
tool.execute.after	Post-tool execution
file.edited	File modification

---

File Structure

.opencode/
|-- index.ts              # Plugin entry
|-- opencode.json         # Config
|-- agents/               # Agent definitions (with embedded prompts)
|-- commands/            # Command files
|-- instructions/        # Rules
|-- skills/              # Domain skills
|-- tools/               # Custom tools

---

Requirements

• OpenCode 1.8.0+
• Node.js 18+

---

License

MIT License