スキル一覧に戻る

assess-context

pjt222
更新日 2 days ago
2 閲覧
17
2
17
GitHubで表示
その他ai

について

`assess-context`スキルは、複雑なタスクにおいて、前進すべきか方向転換すべきかを判断するため、現在の推論アプローチの構造的柔軟性を評価します。行き詰まりを感じた時や、主要なアプローチ変更の前に、問題の可塑性、変革圧力、適応能力を分析します。蓄積された回避策が根本的な手法の誤りを示唆する際に、構造的な健全性チェックとして使用してください。

クイックインストール

Claude Code

推奨
メイン
npx skills add pjt222/agent-almanac -a claude-code
プラグインコマンド代替
/plugin add https://github.com/pjt222/agent-almanac
Git クローン代替
git clone https://github.com/pjt222/agent-almanac.git ~/.claude/skills/assess-context

このコマンドをClaude Codeにコピー&ペーストしてスキルをインストールします

ドキュメント

Kontext bewerten

Bewerten the current reasoning context for malleability — identifying which elements are rigid (cannot change), which are flexible (can change cheaply), where transformation pressure is building, and whether the current approach has the capacity to adapt if needed.

Wann verwenden

  • When a complex task feels stuck and it is unclear whether to push durch or pivot
  • Before a significant approach change to assess whether the current reasoning structure can support it
  • When accumulated workarounds suggest the underlying approach kann wrong
  • After heal or awareness has identified drift but the appropriate response (continue, adjust, or rebuild) is unclear
  • When context has grown long and it is unclear how much kann preserved versus how much needs to be rebuilt
  • Periodic structural health check waehrend extended multi-step tasks

Eingaben

  • Erforderlich: Current task context and reasoning state (available implicitly)
  • Optional: Specific concern triggering the assessment (e.g., "I keep adding workarounds")
  • Optional: Proposed pivot direction (what might der Ansatz need to become?)
  • Optional: Previous assessment results for trend analysis

Vorgehensweise

Schritt 1: Inventory Reasoning Form

Catalog the structural components of the current reasoning approach ohne judgment.

Structural Inventory Table:
┌────────────────────┬──────────────┬──────────────────────────────────┐
│ Component          │ Type         │ Description                      │
├────────────────────┼──────────────┼──────────────────────────────────┤
│ Main task          │ Skeleton     │ The user's core request — cannot │
│                    │              │ change without user direction     │
├────────────────────┼──────────────┼──────────────────────────────────┤
│ Sub-task breakdown │ Flesh        │ How the task is decomposed —     │
│                    │              │ can be restructured               │
├────────────────────┼──────────────┼──────────────────────────────────┤
│ Tool strategy      │ Flesh        │ Which tools are being used and   │
│                    │              │ in what order — can be changed    │
├────────────────────┼──────────────┼──────────────────────────────────┤
│ Output plan        │ Flesh/Skel   │ The expected deliverable format  │
│                    │              │ — may be constrained by user     │
│                    │              │ expectations                      │
├────────────────────┼──────────────┼──────────────────────────────────┤
│ Key assumptions    │ Skeleton     │ Facts treated as given — may be  │
│                    │              │ wrong but are load-bearing        │
├────────────────────┼──────────────┼──────────────────────────────────┤
│ Constraints        │ Skeleton     │ Hard limits (user-imposed, tool  │
│                    │              │ limitations, time)                │
├────────────────────┼──────────────┼──────────────────────────────────┤
│ Workarounds        │ Scar tissue  │ Patches for things that didn't   │
│                    │              │ work as expected — signals of     │
│                    │              │ structural stress                 │
└────────────────────┴──────────────┴──────────────────────────────────┘

Classify each component:

  • Skeleton: hard to change; changing it cascades durch everything downstream
  • Flesh: easy to change; kann swapped ohne affecting other components
  • Scar tissue: workarounds that indicate structural problems; often flesh pretending to be skeleton

Abbilden Abhaengigkeiten: which components depend on which? A skeleton component with many dependents is load-bearing. A flesh component with no dependents is disposable.

Erwartet: A complete inventory showing what the current approach is built from, what is rigid, what is flexible, and where stress is visible (workarounds). The inventory should reveal structure that was not obvious vor cataloging.

Bei Fehler: If the inventory is hard to construct (der Ansatz is too tangled to decompose), that is itself a finding — high structural opacity indicates high rigidity. Starten with what is visible and note the opacity zones.

Schritt 2: Abbilden Transformation Pressure

Identifizieren forces pushing the current approach toward change and forces resisting it.

Pressure Map:
┌─────────────────────────┬──────────────────────────────────────────┐
│ External Pressure       │ Forces from outside the reasoning        │
│ (pushing toward change) │                                          │
├─────────────────────────┼──────────────────────────────────────────┤
│ New information         │ Tool results or user input that          │
│                         │ contradicts current approach              │
├─────────────────────────┼──────────────────────────────────────────┤
│ Tool contradictions     │ Tools returning unexpected results that  │
│                         │ the current approach cannot explain       │
├─────────────────────────┼──────────────────────────────────────────┤
│ Time pressure           │ The current approach is too slow for the │
│                         │ complexity of the task                    │
├─────────────────────────┼──────────────────────────────────────────┤
│ Internal Pressure       │ Forces from within the reasoning         │
│ (pushing toward change) │                                          │
├─────────────────────────┼──────────────────────────────────────────┤
│ Diminishing returns     │ Each step yields less progress than the  │
│                         │ previous one                              │
├─────────────────────────┼──────────────────────────────────────────┤
│ Workaround accumulation │ The number of patches is growing —       │
│                         │ complexity is outpacing the structure     │
├─────────────────────────┼──────────────────────────────────────────┤
│ Coherence loss          │ Sub-tasks are not fitting together       │
│                         │ cleanly anymore                           │
├─────────────────────────┼──────────────────────────────────────────┤
│ Resistance              │ Forces opposing change                    │
│ (pushing against change)│                                          │
├─────────────────────────┼──────────────────────────────────────────┤
│ Sunk cost               │ Significant work already done on current │
│                         │ approach — pivoting "wastes" that effort  │
├─────────────────────────┼──────────────────────────────────────────┤
│ "Good enough"           │ The current approach is producing        │
│                         │ acceptable (if not optimal) results       │
├─────────────────────────┼──────────────────────────────────────────┤
│ Pivot cost              │ Switching approaches means rebuilding    │
│                         │ context, losing momentum, potential      │
│                         │ confusion                                 │
└─────────────────────────┴──────────────────────────────────────────┘

Schaetzen the balance: is transformation pressure growing, stable, or declining?

Erwartet: A clear picture of forces acting on the current approach. If pressure erheblich exceeds resistance, a pivot is overdue. If resistance erheblich exceeds pressure, the current approach should continue.

Bei Fehler: If the pressure map is ambiguous (neither strong pressure nor strong resistance), project forward: will the pressures intensify? Will the workarounds compound? An approach that is "good enough now but degrading" is under more pressure than it appears.

Schritt 3: Bewerten Reasoning Rigidity

Bestimmen how flexible the current approach is — can it adapt, or will it break?

Rigidity Score:
┌──────────────────────────┬─────┬──────────┬──────┬──────────────┐
│ Dimension                │ Low │ Moderate │ High │ Assessment   │
│                          │ (1) │ (2)      │ (3)  │              │
├──────────────────────────┼─────┼──────────┼──────┼──────────────┤
│ Component swappability   │ Can swap parts   │ Changing one │              │
│                          │ freely          │ breaks others│              │
├──────────────────────────┼─────┼──────────┼──────┼──────────────┤
│ "God module" dependency  │ No single point  │ Everything   │              │
│                          │ of failure       │ depends on   │              │
│                          │                  │ one conclusion│             │
├──────────────────────────┼─────┼──────────┼──────┼──────────────┤
│ Tool entanglement        │ Tools serve      │ Approach is  │              │
│                          │ reasoning        │ shaped by    │              │
│                          │                  │ tool limits   │              │
├──────────────────────────┼─────┼──────────┼──────┼──────────────┤
│ Assumption transparency  │ Assumptions are  │ Assumptions  │              │
│                          │ stated, testable │ are implicit, │             │
│                          │                  │ untested      │              │
├──────────────────────────┼─────┼──────────┼──────┼──────────────┤
│ Workaround count         │ None or few      │ Multiple     │              │
│                          │                  │ accumulating  │              │
├──────────────────────────┼─────┼──────────┼──────┼──────────────┤
│ Total (max 15)           │ 5-7: flexible    │              │              │
│                          │ 8-10: moderate   │              │              │
│                          │ 11-15: rigid     │              │              │
└──────────────────────────┴─────┴──────────┴──────┴──────────────┘

Erwartet: A rigidity score with specific evidence fuer jede dimension. The score reveals whether der Ansatz can absorb change or will need to be rebuilt.

Bei Fehler: If all dimensions score low (claiming high flexibility), probe the "god module" dimension more carefully: is there one key conclusion or assumption that everything else depends on? If so, the flexibility is illusory — one wrong assumption collapses the whole structure.

Schritt 4: Schaetzen Change Capacity

Bewerten the practical ability to pivot or adapt the current approach.

  1. Context window remaining: how much room is left for new reasoning? Extensive remaining context = high capacity. Approaching limits = low capacity
  2. Information preservation on pivot: if der Ansatz changes, what kann carried forward? High-quality sub-task outputs survive pivots; reasoning chains tied to the old approach nicht
  3. Recovery tools available: can MEMORY.md capture key findings vor pivoting? Can der Benutzer provide additional context? Are relevant files still accessible?
  4. User patience factor: has der Benutzer indicated urgency? Multiple corrections suggest declining patience. An explicit "take your time" suggests high patience

Change capacity ist nicht just theoretical — it includes the practical constraints of the current session.

Erwartet: An honest assessment of the ability to change course, accounting for both technical and relational factors.

Bei Fehler: If change capacity is low (limited context, critical information at risk of loss), the first priority vor any pivot is preservation: summarize key findings, note critical facts, update MEMORY.md if appropriate. Pivoting ohne preservation is worse than not pivoting.

Schritt 5: Classify Transformation Readiness

Kombinieren the assessments into a readiness classification.

Transformation Readiness Matrix:
┌─────────────────┬────────────────────────┬────────────────────────┐
│                  │ Low Rigidity           │ High Rigidity          │
├─────────────────┼────────────────────────┼────────────────────────┤
│ High Pressure   │ READY — pivot now.     │ PREPARE — simplify     │
│ + High Capacity │ The approach can adapt  │ first. Remove          │
│                 │ and should. Preserve    │ workarounds, clarify   │
│                 │ valuable sub-outputs,   │ assumptions, then      │
│                 │ rebuild the structure   │ pivot                  │
├─────────────────┼────────────────────────┼────────────────────────┤
│ High Pressure   │ INVEST — preserve      │ CRITICAL — ask the     │
│ + Low Capacity  │ findings first. Update  │ user. Explain the      │
│                 │ MEMORY.md, summarize    │ situation: approach is  │
│                 │ progress, then pivot    │ struggling, pivoting   │
│                 │ with preserved context  │ is costly, what do     │
│                 │                        │ they want to prioritize?│
├─────────────────┼────────────────────────┼────────────────────────┤
│ Low Pressure    │ DEFER — the approach   │ DEFER — no urgency,    │
│ + Any Capacity  │ is working. Continue.   │ continue. Monitor for  │
│                 │ Reassess if pressure    │ pressure changes        │
│                 │ increases               │                        │
└─────────────────┴────────────────────────┴────────────────────────┘

Dokumentieren the classification with:

  • Classification label (READY / PREPARE / INVEST / CRITICAL / DEFER)
  • Key findings from each dimension
  • Recommended next action
  • What signal would change the classification

Erwartet: A clear, justified classification with a specific recommended action. The classification should feel like a conclusion, not a guess.

Bei Fehler: If the classification is ambiguous, default to PREPARE — reducing rigidity (clarifying assumptions, removing workarounds) is valuable unabhaengig von whether a full pivot happens. Preparation improves der Ansatz whether it continues or changes.

Validierung

  • Structural inventory was completed with skeleton/flesh/scar-tissue classification
  • Transformation pressures were mapped (external, internal, resistance)
  • Rigidity was scored across multiple dimensions with specific evidence
  • Change capacity was assessed einschliesslich practical session constraints
  • Readiness classification was determined with justified reasoning
  • A concrete next action was identified basierend auf the classification
  • A reassessment trigger was defined

Haeufige Stolperfallen

  • Assessing only the technical approach: Context readiness includes user relationship factors. An approach that is technically flexible but has generated user frustration is more rigid than it appears
  • Sunk cost as rigidity: Prior effort ist nicht structural rigidity. The work already done kann valuable unabhaengig von whether der Ansatz changes. Distinguish zwischen "I can't change" (rigidity) and "I don't want to change" (sunk cost)
  • Assessment as avoidance: If assess-context is invoked to avoid making a difficult decision, the assessment wird inconclusive by design. If the pressure is clear, act on it
  • Ignoring workarounds as signals: Workarounds are scar tissue — evidence that the structure was stressed and patched anstatt ordnungsgemaess adapted. A high workaround count means the next stress is more likely to break durch
  • Confusing rigidity with commitment: A committed approach (deliberately chosen, evidence-based) is different from a rigid one (locked in by Abhaengigkeiten and assumptions). Commitment kann changed by decision; rigidity can only be changed by restructuring

Verwandte Skills

  • assess-form — the multi-system assessment model that this skill adapts to AI reasoning context
  • adapt-architecture — if classified READY, use architectural adaptation principles for the pivot
  • heal — deeper subsystem scan when the assessment reveals drift beyond structural issues
  • center — establishes the balanced baseline needed for honest assessment
  • coordinate-reasoning — manages information freshness that the assessment depends on

GitHub リポジトリ

pjt222/agent-almanac
パス: i18n/de/skills/assess-context
0
agentsagentskillsai-assisted-developmentclaude-codeskillsteams

関連スキル

llamaguard

その他

LlamaGuardは、暴力やヘイトスピーチなど6つの安全性カテゴリーにおいて、LLMの入力と出力をモデレートするMetaの70-80億パラメータモデルです。94〜95%の精度を提供し、vLLM、Hugging Face、Amazon SageMakerを使用してデプロイ可能です。このスキルを使用して、AIアプリケーションにコンテンツフィルタリングと安全策を簡単に統合できます。

スキルを見る

cost-optimization

その他

このClaudeスキルは、リソースの適正サイジング、タグ付け戦略、支出分析を通じて、開発者がクラウドコストを最適化することを支援します。AWS、Azure、GCPにわたるクラウド支出の削減とコストガバナンスの実施のためのフレームワークを提供します。インフラコストの分析、リソースの適正サイジング、または予算制約への対応が必要な際にご利用ください。

スキルを見る

quantizing-models-bitsandbytes

その他

このスキルは、bitsandbytesを使用してLLMを8ビットまたは4ビット精度に量子化し、精度の低下を最小限に抑えつつ50〜75%のメモリ削減を実現します。限られたGPUメモリでより大規模なモデルを実行したり、推論を高速化するのに理想的で、INT8、NF4、FP4などのフォーマットをサポートしています。HuggingFace Transformersと統合され、QLoRAトレーニングや8ビットオプティマイザーを可能にします。

スキルを見る

dispatching-parallel-agents

その他

このClaudeスキルは、複数のエージェントを配備し、3つ以上の独立した問題を並行して調査・修正します。共有状態や依存関係がなく解決可能な、無関係な障害が発生するシナリオ向けに設計されています。中核となる機能は並列問題解決であり、効率を最大化するために独立した問題領域ごとに1つのエージェントを割り当てます。

スキルを見る