Skip to content

johnfkoo951/cmdspace-plugins / plugins/dev-orchestrator/skills/focus

Deep focus mode using a single high-tier agent for complex problems requiring sustained reasoning. When triggered by '/focus', 'focus mode', '집중 모드', or '깊이 생각해줘', this skill dedicates a single Opus-tier agent to one problem without context switching, multi-agent coordination, or parallelization.

Focus — Single Agent Deep Reasoning Mode

Focus mode applies concentrated, deep reasoning to a single problem. No multi-agent coordination, no context switching — just one powerful agent thinking deeply.

When to Use

  • Architecture decisions requiring careful trade-off analysis
  • Complex debugging that needs systematic root cause analysis
  • Code that requires understanding deep domain logic
  • Decisions with significant long-term consequences

Workflow

Step 1: Understand the Problem

Use AskUserQuestion to clarify:

{
  "questions": [
    {
      "question": "What kind of deep analysis do you need?",
      "header": "Focus Area",
      "options": [
        {"label": "Architecture decision", "description": "Evaluate trade-offs and design a solution. Uses architect agent."},
        {"label": "Critical review", "description": "Find flaws and risks in existing design/code. Uses critic agent."},
        {"label": "Complex debugging", "description": "Systematic root cause analysis. Uses executor agent with Opus override."},
        {"label": "Security analysis", "description": "Deep security vulnerability assessment. Uses security-reviewer agent."}
      ],
      "multiSelect": false
    }
  ]
}

Step 2: Set Up Context

Before dispatching the focused agent:

  1. Read all relevant files to build comprehensive context
  2. Gather any related documentation, ADRs, or specifications
  3. Identify constraints and requirements

Step 3: Dispatch Single Agent

Based on the focus area, dispatch ONE agent with Opus tier:

  • Architecture decision → architect (Opus)
  • Critical review → critic (Opus)
  • Complex debugging → executor with Opus override
  • Security analysis → security-reviewer (Opus)

Provide the agent with:

  • Complete context (all relevant files already read)
  • Clear problem statement
  • Specific questions to answer or decisions to make
  • Constraints and requirements

Step 4: Present Results

The focused agent's output should be presented as-is, without additional agent processing. The value of focus mode is the depth of a single agent's analysis, not a pipeline.

Key Principle

Less is more. One agent thinking deeply about one problem produces better results than five agents skimming five aspects.

Differences from Other Modes

AspectFocusAutopilotPipeline
Agents1ManySequential chain
TierAlways OpusMixedMixed
SpeedSlow, thoroughFast, parallelMedium
Best forDecisionsExecutionWorkflows
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 469kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 321kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 321kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 321kUpdated Claude CodeCodexVerifiedInstall