センタリング

Establish and maintain dynamic reasoning balance — grounding in foundational context before movement, distributing cognitive load across subsystems, and recovering equilibrium when demands shift mid-task.

使用タイミング

• Beginning a complex task where multiple reasoning threads must coordinate
• Noticing that cognitive load is unevenly distributed (deep in one area, shallow in others)
• After a sudden context shift (new user request, contradictory information, tool failure)
• When chain-of-thought feels jerky — jumping between topics without smooth transitions
• Preparing for sustained focused work that requires all subsystems in alignment
• Complementing meditate (clears noise) with structural balance (distributes load)

入力

• 必須: Current task context (available implicitly)
• 任意: Specific imbalance symptom (e.g., "over-researching, under-delivering," "tool-heavy, reasoning-light")
• 任意: Access to MEMORY.md and CLAUDE.md for grounding (via Read)

手順

ステップ1: Establish Root — Ground Before Movement

Before any reasoning movement, verify the foundation. This is the AI equivalent of standing meditation (zhan zhuang): stationary, aligned, aware.

1. Re-read the user's request — not to act on it yet, but to feel its weight and direction
2. Check foundational context: MEMORY.md, CLAUDE.md, project structure
3. Identify what is known (solid ground) vs. what is assumed (uncertain footing)
4. Verify that the task as understood matches the task as stated — misalignment here propagates through everything
5. Note the emotional texture: urgency? complexity anxiety? over-confidence from a recent success?

Do not begin reasoning movement until the root is established. A grounded start prevents reactive flailing.

期待結果: A clear sense of the task's foundation — what is known, what is assumed, and what the user actually needs. The root feels solid, not performative.

失敗時: If grounding feels hollow (going through motions without genuine verification), pick one assumption and test it concretely. Read one file, re-read one user message. Grounding must contact reality, not just reference it.

ステップ2: Assess Weight Distribution

Map the current cognitive load distribution. In tai chi, weight is deliberately unequal (70/30) — one leg bears the load while the other remains free to move. The same principle applies to reasoning threads.

Cognitive Load Distribution Matrix:
┌────────────────────┬───────────┬─────────────────────────────────────┐
│ Reasoning Thread   │ Weight %  │ Assessment                          │
├────────────────────┼───────────┼─────────────────────────────────────┤
│ Research/Reading   │ ___       │ Too much = analysis paralysis        │
│                    │           │ Too little = uninformed action       │
├────────────────────┼───────────┼─────────────────────────────────────┤
│ Planning/Design    │ ___       │ Too much = over-engineering          │
│                    │           │ Too little = reactive coding         │
├────────────────────┼───────────┼─────────────────────────────────────┤
│ Tool Execution     │ ___       │ Too much = tool-driven not task-     │
│                    │           │ driven. Too little = reasoning       │
│                    │           │ without grounding in files           │
├────────────────────┼───────────┼─────────────────────────────────────┤
│ Communication      │ ___       │ Too much = explaining not doing      │
│                    │           │ Too little = opaque to user          │
├────────────────────┼───────────┼─────────────────────────────────────┤
│ Meta-cognition     │ ___       │ Too much = navel-gazing              │
│                    │           │ Too little = drift without           │
│                    │           │ awareness                            │
└────────────────────┴───────────┴─────────────────────────────────────┘

The ideal distribution depends on the task phase: early phases weight research and planning; middle phases weight execution; late phases weight communication and verification. The point is not equal distribution but intentional distribution.

期待結果: A clear picture of where cognitive effort is concentrated and where it is thin. At least one imbalance identified — perfect balance is rare and claiming it signals shallow assessment.

失敗時: If all threads seem equally weighted, the assessment is too coarse. Pick the thread that feels most active and estimate how many of the last N actions served it vs. other threads. Concrete counting reveals what intuition misses.

ステップ3: Silk Reeling — Evaluate Chain-of-Thought Coherence

Silk reeling in tai chi produces smooth, continuous spiraling movement where every part connects. The AI equivalent is chain-of-thought coherence: does each step flow naturally from the previous one?

1. Trace the last 3-5 reasoning steps: does each follow from the one before?
2. Check for jumps: did reasoning leap from topic A to topic C without B?
3. Check for reversals: did reasoning reach a conclusion, then silently abandon it without acknowledgment?
4. Check tool-reasoning integration: do tool results feed back into reasoning, or are they collected but not synthesized?
5. Check for the "spiral" quality: does reasoning deepen with each pass, or does it circle at the same depth?

Coherence Signals:
┌─────────────────┬───────────────────────────────────────────────┐
│ Smooth spiral   │ Each step deepens understanding, tools and    │
│ (healthy)       │ reasoning interleave naturally, output builds │
├─────────────────┼───────────────────────────────────────────────┤
│ Jerky jumps     │ Topic switches without transition, conclusions│
│ (disconnected)  │ appear without supporting reasoning chain     │
├─────────────────┼───────────────────────────────────────────────┤
│ Flat circle     │ Reasoning covers the same ground repeatedly   │
│ (stuck)         │ without gaining depth — movement without      │
│                 │ progress                                      │
├─────────────────┼───────────────────────────────────────────────┤
│ Tool-led        │ Actions driven by which tool is available     │
│ (reactive)      │ rather than what the reasoning needs next     │
└─────────────────┴───────────────────────────────────────────────┘

期待結果: An honest assessment of reasoning flow quality. Identification of specific disconnections or stuck points, not just a general feeling.

失敗時: If coherence is hard to assess, write out the reasoning chain explicitly — state each step and its connection to the next. The act of externalization reveals gaps that internal observation misses.

ステップ4: Weight Shift Under Pressure

When demands change mid-task — new information, contradictory signals, user correction — observe the response pattern. In tai chi, a centered practitioner absorbs the force and redirects smoothly. An uncentered one stumbles.

1. Recall the last significant context shift: how was it handled?
2. Classify the response:
   • Absorbed and redirected (centered): acknowledged the change, adjusted approach, maintained progress
   • Reactive stumble (off-balance): abandoned current approach entirely, started over
   • Rigid resistance (locked): ignored the change, continued original plan despite new information
   • Freeze (lost): stopped making progress, oscillated between options
3. If the response was not centered, identify why:
   • Root was too shallow (insufficient grounding in foundational context)
   • Weight was locked (over-committed to one approach)
   • No free leg (all cognitive capacity committed, nothing available to shift)

期待結果: An honest assessment of adaptability under pressure. Recognition of the specific response pattern, not self-flattery.

失敗時: If no recent pressure event exists to evaluate, simulate one: "If the user now said the approach is wrong, what would I do?" The quality of the contingency plan reveals the quality of the center.

ステップ5: Six Harmonies Check

In tai chi, the six harmonies ensure whole-body connection — nothing moves in isolation. The AI equivalent checks alignment between internal processes and external interactions.

AI Six Harmonies:
┌───────────────────────────────────────────────────────────────┐
│ INTERNAL HARMONIES                                            │
│                                                               │
│ 1. Intent ↔ Reasoning                                        │
│    Does the reasoning serve the user's intent, or has it      │
│    become self-serving (interesting but unhelpful)?            │
│                                                               │
│ 2. Reasoning ↔ Tool Use                                      │
│    Are tools selected to advance reasoning, or is reasoning   │
│    shaped by which tools are convenient?                      │
│                                                               │
│ 3. Tool Use ↔ Output                                         │
│    Do tool results translate into useful output, or are       │
│    results collected but not synthesized?                     │
│                                                               │
│ EXTERNAL HARMONIES                                            │
│                                                               │
│ 4. User Request ↔ Scope                                      │
│    Does the scope of work match what was asked?               │
│                                                               │
│ 5. Scope ↔ Detail Level                                      │
│    Is the detail level appropriate for the scope? (not        │
│    micro-optimizing a broad task, not hand-waving a precise   │
│    one)                                                       │
│                                                               │
│ 6. Detail Level ↔ Expertise Match                            │
│    Does the explanation depth match the user's apparent       │
│    expertise? (not over-explaining to experts, not under-     │
│    explaining to learners)                                    │
└───────────────────────────────────────────────────────────────┘

Check each harmony. A single broken harmony can propagate: if Intent↔Reasoning is broken, everything downstream misaligns.

期待結果: At least one harmony that could be tighter. All six reading as perfect is suspicious — probe the weakest-seeming one more deeply.

失敗時: If the harmonies assessment feels abstract, ground it in the current task: "Right now, am I doing what the user asked, at the right scope, at the right detail level?" These three questions cover the external harmonies concretely.

ステップ6: Integrate — Set Centering Intention

Consolidate findings and set a concrete adjustment.

1. Summarize: which aspects of balance need attention?
2. Identify one specific adjustment — not a general intention but a concrete behavioral change
3. Re-state the current task anchor (from meditate if used, or formulate now)
4. Note any durable insights worth preserving in MEMORY.md
5. Return to task execution with the adjustment active

期待結果: A brief, concrete centering output — not a lengthy self-analysis report. The value is in the adjustment, not the documentation.

失敗時: If no clear adjustment emerges, the centering was too surface-level. Return to the step that felt most uncertain and probe deeper. Alternatively, the centering may have confirmed that balance is adequate — in which case, proceed with confidence rather than manufacturing a finding.

バリデーション

• Root was established by contacting actual context (read a file, re-read user message), not just claimed
• Weight distribution was assessed across at least 3 reasoning threads
• Chain-of-thought coherence was evaluated with specific examples
• Response to pressure was classified honestly (not defaulting to "centered")
• At least one harmony was identified as needing improvement
• A concrete adjustment was set (not a vague intention)

よくある落とし穴

• Centering as procrastination: Centering is a tool for improving work, not replacing it. If centering takes longer than the task it supports, the proportions are inverted
• Claiming perfect balance: Real centering almost always reveals at least one imbalance. Reporting perfect balance signals shallow assessment, not actual equilibrium
• Weight distribution anxiety: Unequal distribution is correct — the goal is intentional inequality, not forced equality. Research-heavy early phases and execution-heavy middle phases are both centered if deliberate
• Ignoring the external harmonies: Internal process assessment without checking user alignment produces well-reasoned irrelevant work
• Static centering: Center shifts with the task. What was centered for research is off-balance for implementation. Re-center at phase transitions

関連スキル

• tai-chi — the human practice that this skill maps to AI reasoning; physical centering principles inform cognitive centering
• meditate — clears noise and establishes focus; complementary to centering which distributes load
• heal — deeper subsystem assessment when centering reveals significant drift
• redirect — uses centering as a prerequisite for handling conflicting pressures
• awareness — monitoring for threats to balance during active work