Portable multi-agent harness for .agents-based skills, workflows, and standards-aware agent teams across Antigravity, Claude Code, Codex, Cursor, OpenCode, and more.
Ever wished your AI assistant had coworkers? That's what oh-my-agent does.
Instead of one AI doing everything (and getting confused halfway through), oh-my-agent splits work across specialized agents — frontend, backend, architecture, QA, PM, DB, mobile, infra, debug, design, and more. Each one knows its domain deeply, has its own tools and checklists, and stays in its lane.
Works with all major AI IDEs: Antigravity, Claude Code, Cursor, Gemini CLI, Codex CLI, OpenCode, and more.
The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
When a workflow resolves an agent to the same vendor as the current runtime, it should use that vendor's native subagent path first. Cross-vendor tasks fall back to oma agent:spawn.
Quick Start
# macOS / Linux — auto-installs bun & uv if missing
curl -fsSL https://raw.githubusercontent.com/first-fluke/oh-my-agent/main/cli/install.sh | bash
# Windows (PowerShell) — auto-installs bun & uv if missing
irm https://raw.githubusercontent.com/first-fluke/oh-my-agent/main/cli/install.ps1 | iex
# Or manual (any OS, requires bun + uv)
bunx oh-my-agent@latest
Install via Agent Package Manager
Microsoft's Agent Package Manager (APM) — not to be confused with oma-observability's APM (Application Performance Monitoring).
# 22 skills, deployed to every detected runtime
# (.claude, .cursor, .codex, .opencode, .github, .agents)
apm install first-fluke/oh-my-agent
# A single skill
apm install first-fluke/oh-my-agent/.agents/skills/oma-frontend
APM reads .claude-plugin/plugin.json's skills: .agents/skills/ pointer, so the .agents/ SSOT is the only source — no build step or mirror.
APM ships the 22 skills only. For workflows, rules, oma-config.yaml, keyword-detection hooks, and the oma agent:spawn CLI, use bunx oh-my-agent@latest. Pick one distribution per project to avoid drift.
| Agent | What They Do |
|-------|-------------|
| oma-architecture | Architectural tradeoffs, boundaries, ADR/ATAM/CBAM-aware analysis |
| oma-backend | APIs in Python, Node.js, or Rust |
| oma-brainstorm | Explores ideas before you commit to building |
| oma-coordination | Manual step-by-step multi-agent coordination guide |
| oma-db | Schema design, migrations, indexing, vector DB |
| oma-debug | Root cause analysis, fixes, regression tests |
| oma-design | Design systems, tokens, accessibility, responsive |
| oma-dev-workflow | CI/CD, releases, monorepo automation |
| oma-frontend | React/Next.js, TypeScript, Tailwind CSS v4, shadcn/ui |
| oma-hwp | HWP/HWPX/HWPML to Markdown conversion |
| oma-image | Multi-vendor AI image generation |
| oma-mobile | Flutter cross-platform apps |
| oma-observability | Observability router — APM/RUM, metrics/logs/traces/profiles, SLO, incident forensics, transport tuning |
| oma-orchestrator | Parallel agent execution via CLI |
| oma-pdf | PDF to Markdown conversion |
| oma-pm | Plans tasks, breaks down requirements, defines API contracts |
| oma-qa | OWASP security, performance, accessibility review |
| oma-recap | Conversation history recap and themed work summaries |
| oma-scm | SCM (software configuration management) — branching, merges, worktrees, baselines; Conventional Commits |
| oma-search | Intent-based search router with trust scoring — docs, web, code, local |
| oma-tf-infra | Multi-cloud Terraform IaC (Infrastructure as Code) |
| oma-translator | Natural multilingual translation |
How It Works
Just chat. Describe what you want and oh-my-agent figures out which agents to use.
You: "Build a TODO app with user authentication"
→ PM plans the work
→ Backend builds auth API
→ Frontend builds React UI
→ DB designs schema
→ QA reviews everything
→ Done: coordinated, reviewed code
Auto-detection: You don't even need slash commands — keywords like "architecture", "plan", "review", and "debug" in your message (in 11 languages!) auto-activate the right workflow.
CLI
# Install globally
bun install --global oh-my-agent # or: brew install oh-my-agent
# Use anywhere
oma doctor # Health check
oma dashboard # Real-time agent monitoring
oma link # Regenerate .claude/.codex/.gemini/etc. from .agents/
oma agent:spawn backend "Build auth API" session-01
oma agent:parallel -i backend:"Auth API" frontend:"Login form"
oma retro 7d --compare # Engineering retrospective with metrics + trends
oma recap --window 1d # Cross-tool conversation history recap
oma search fetch <url> # Mechanical search with auto-escalating strategies
oma image generate "cat" # Multi-vendor AI image generation
oma export cursor # Project skills as `.cursor/rules` for external IDEs
Model selection follows two layers:
Same-vendor native dispatch uses the generated vendor agent definition in .claude/agents/, .codex/agents/, or .gemini/agents/.
Cross-vendor or fallback CLI dispatch uses the vendor defaults in .agents/skills/oma-orchestrator/config/cli-config.yaml.
Per-Agent Models
Set model_preset in .agents/oma-config.yaml to choose which AI models each agent uses:
A Claude Code plugin that automatically captures everything Claude does during your coding sessions, compresses it with AI (using Claude's agent-sdk), and injects relevant context back into future sessions.
