Built-in Skills

Built-in skills ship with every OpenCode + oMO installation. No setup, no configuration, no install step. They're available from your first session.

git-master

Trigger: Any git operation — commits, rebases, squash, history search, blame, bisect.

git-master wraps all git operations with atomic commit discipline, safe rebase/squash workflows, and history-aware search. It prevents common git mistakes (force-pushing to main, losing uncommitted work) and structures commits for readability.

When to use it:

• Committing changes with meaningful messages
• Interactive rebase workflows (but not via -i — it uses non-interactive equivalents)
• Finding when something was introduced (git log -S, git blame)
• Squashing commits safely

// Sisyphus automatically loads git-master for git operations
task(category="quick", load_skills=["git-master"], prompt="Commit the auth changes")

review-work

Trigger: After completing any significant implementation. Say "review my work", "review changes", "QA my work", or "verify implementation".

review-work launches 5 parallel background sub-agents, each with a specific review lens:

Sub-agent	Role
Oracle (goal/constraint)	Verifies implementation matches stated goals
Oracle (code quality)	Checks for code smells, dead code, poor patterns
Oracle (security)	Scans for vulnerabilities, secret leaks, unsafe patterns
Hands-on QA	Actually runs tests, checks builds, verifies behavior
Context miner	Gathers context from GitHub, git history, Slack, Notion

All 5 must pass for the review to pass. If any fails, the review surfaces the specific issue.

// After implementing a feature
task(category="unspecified-high", load_skills=["review-work"], prompt="Review the auth module changes")

ai-slop-remover

Trigger: When you want to clean up AI-generated code smells. Operates on a single file at a time — for multiple files, run in parallel.

ai-slop-remover identifies and removes common AI-generated code patterns while preserving functionality:

• Unnecessary comments that restate what the code does
• Overly defensive null checks that the type system already handles
• Redundant variable declarations
• Overly verbose error handling that obscures the happy path
• Generic naming that adds no clarity

// Clean up a specific file
task(category="quick", load_skills=["ai-slop-remover"], prompt="Remove AI slop from src/auth/handlers.ts")

frontend-ui-ux

Trigger: Building any web UI — components, pages, layouts, dashboards, landing pages.

frontend-ui-ux applies design-first principles to frontend work. It enforces:

• Metric-based design rules (not vibes)
• Strict component architecture
• CSS hardware acceleration
• Balanced design engineering (no over-engineering, no under-styling)

This skill overrides the default LLM tendency toward generic, template-looking UI.

// Build a dashboard component
task(category="visual-engineering", load_skills=["frontend-ui-ux"], prompt="Build a user analytics dashboard with charts")

Disabling Built-in Skills

If a built-in skill conflicts with your workflow, disable it in your oMO config:

{
  "disabled_skills": ["playwright", "agent-browser"]
}

Note: playwright and agent-browser are also built-in skills that can be disabled. The disabled_skills array accepts any skill name — built-in or plugin.