MCP HubMCP Hub
스킬 목록으로 돌아가기

dissolve-form

pjt222
업데이트됨 2 days ago
1 조회
17
2
17
GitHub에서 보기
기타general

정보

dissolve-form 스킬은 핵심 기능을 보존하면서 경직된 시스템 구조를 통제적으로 해체합니다. 이 스킬은 경직성 매핑, 해체 순서 설계, 그리고 점진적 변경이 불가능할 정도로 경화된 시스템의 기술 부채를 안전하게 분해하는 기능을 포함합니다. 기술 부채가 모든 진전을 막거나 시스템을 재구성하기 전에 유연화가 필요할 때 사용하세요.

빠른 설치

Claude Code

추천
기본
npx skills add pjt222/agent-almanac -a claude-code
플러그인 명령대체
/plugin add https://github.com/pjt222/agent-almanac
Git 클론대체
git clone https://github.com/pjt222/agent-almanac.git ~/.claude/skills/dissolve-form

Claude Code에서 이 명령을 복사하여 붙여넣어 스킬을 설치하세요

문서

Form aufloesen

Perform controlled dismantling of rigid system structures — dissolving calcified architecture, accumulated technical debt, and organizational rigidity while preserving the essential capabilities ("imaginal discs") that will seed the new form.

Wann verwenden

  • Form assessment (see assess-form) classified das System as PREPARE or CRITICAL (too rigid to transform directly)
  • A system is so calcified that incremental change is impossible
  • Technical debt has compounded to the point where it blocks all forward progress
  • An organizational structure has become so rigid that it can't adapt to new requirements
  • Before adapt-architecture when the current form muss softened vor it kann reshaped
  • Legacy system decommissioning where value muss extracted vor shutdown

Eingaben

  • Erforderlich: Form assessment showing high rigidity (from assess-form)
  • Erforderlich: Identification of essential capabilities to preserve (imaginal discs)
  • Optional: Target form (what should emerge nach dissolution; kann unknown)
  • Optional: Dissolution timeline and constraints
  • Optional: Stakeholder concerns about specific components
  • Optional: Previous dissolution attempts and their outcomes

Vorgehensweise

Schritt 1: Identifizieren Imaginal Discs

In biological metamorphosis, imaginal discs are clusters of cells innerhalb the caterpillar that survive dissolution and become the butterfly's organs. Identifizieren the essential capabilities that must survive.

  1. Catalog every capability the current system provides:
    • User-facing features
    • Data processing functions
    • Integration points with external systems
    • Institutional knowledge embedded in the code/process
    • Business rules (often implicit, undocumented)
  2. Classify each capability:
    • Imaginal disc (must survive): core business logic, critical integrations, irreplaceable data
    • Replaceable tissue (kann rebuilt): UI, infrastructure, standard algorithms
    • Dead tissue (should not survive): workarounds for bugs that no longer exist, compatibility shims for dead systems, features nobody uses
  3. Extrahieren imaginal discs into portable form:
    • Dokumentieren business rules explicitly (they may only exist as code comments or tribal knowledge)
    • Extrahieren critical algorithms into standalone, tested modules
    • Exportieren essential data in format-independent representations
    • Erfassen integration contracts and their actual (not documented) behavior

Erwartet: A clear inventory of capabilities classified as essential (preserve), replaceable (rebuild), or dead (discard). Essential capabilities are extracted into portable form vor dissolution begins.

Bei Fehler: If imaginal disc identification is uncertain (stakeholders disagree on what's essential), err on the side of preservation. Extrahieren more capabilities than you think you'll need — discarding nach dissolution is easy; recovering lost knowledge is often impossible.

Schritt 2: Abbilden Dissolution Sequence

Bestimmen the order in which structural elements wird dissolved — outer layers first, core last.

  1. Order by Abhaengigkeit depth:
    • Layer 1 (outermost): components with no dependents — nothing breaks when they're removed
    • Layer 2: components whose dependents are only Layer 1 items (already dissolved)
    • Layer 3: components with deeper Abhaengigkeiten — removing these requires careful interface management
    • Layer N (core): load-bearing components that everything depends on — dissolved last
  2. Fuer jede layer, define:
    • What is dissolved (removed, decommissioned, archived)
    • What replaces it (new component, nothing, or temporary stub)
    • What interfaces muss maintained for the remaining layers
    • How to verify das System still functions nach this layer is dissolved
  3. Erstellen dissolution checkpoints:
    • After each layer, the remaining system muss tested and verified operational
    • Each checkpoint is a stable state from which dissolution can pause
    • If a layer's dissolution causes unexpected breakage, restore from the previous checkpoint
Dissolution Sequence (outside in):
┌─────────────────────────────────────────────────────────────────┐
│ Layer 1: Dead features, unused integrations, orphaned code      │
│          → Remove. Nothing depends on these.                    │
│                                                                 │
│ Layer 2: Replaceable UI, standard infrastructure                │
│          → Replace with modern equivalents or stubs             │
│                                                                 │
│ Layer 3: Business logic wrappers, data access layers            │
│          → Extract imaginal discs, then dissolve                │
│                                                                 │
│ Layer 4 (core): Load-bearing structures, data stores            │
│          → Dissolve last, with full replacement ready           │
└─────────────────────────────────────────────────────────────────┘

Erwartet: A layer-ordered dissolution sequence where each step is safe (checkpoint verified) and reversible (previous checkpoint is restorable). The most critical components are dissolved last when das Team has the most experience and confidence.

Bei Fehler: If Abhaengigkeit mapping reveals circular Abhaengigkeiten (A depends on B depends on A), these cycles muss broken vor sequenced dissolution is possible. Introduce an interface zwischen A and B, break the cycle, then proceed with the sequence.

Schritt 3: Perform Interface Archaeology

Before dissolving rigid structures, excavate and document their actual interfaces — not what's documented, but what's actually in use.

  1. Instrument current interfaces:
    • Log every call, message, or data exchange at each interface
    • Ausfuehren for mindestens one full business cycle (daily, weekly, monthly — whatever is relevant)
    • Capture actual payload shapes, not just documented schemas
  2. Vergleichen actual vs. documented behavior:
    • What documented interfaces are never called? (candidates for Layer 1 dissolution)
    • What undocumented interfaces are actively used? (hidden Abhaengigkeiten — muss preserved or explicitly replaced)
    • What Grenzfaelle does the actual traffic reveal that documentation doesn't mention?
  3. Erstellen an interface contract from actual behavior:
    • This contract becomes the specification for any replacement
    • Einschliessen real examples of inputs and outputs
    • Dokumentieren Fehlerbehandlung behavior (what actually happens, not what should happen)

Erwartet: An empirically-derived interface contract that accurately represents how das System actually communicates, einschliesslich undocumented behaviors and hidden Abhaengigkeiten.

Bei Fehler: If instrumentation is too invasive (impacts performance or requires code changes), sample traffic stattdessen of capturing everything. If the business cycle is too long to wait, use the available data supplemented by stakeholder interviews about "what calls what in which situations."

Schritt 4: Ausfuehren Controlled Dissolution

Systematically remove structural elements while maintaining imaginal disc viability.

  1. Beginnen with Layer 1 (outermost, no dependents):
    • Entfernen dead features and unused code
    • Archive (don't delete) for reference
    • Verify: system still passes all tests, no runtime errors
  2. Fortschreiten durch each layer:
    • Fuer jede component being dissolved: a. Verifizieren imaginal discs wurden extracted (Step 1) b. Installieren replacement or stub (if dependents remain) c. Entfernen die Komponente d. Ausfuehren validation suite e. Ueberwachen for unexpected Seiteneffekts
    • At each checkpoint: document the current system state, verify operational status
  3. Behandeln dissolution resistance:
    • Some components resist dissolution (hidden Abhaengigkeiten surface)
    • When a removal causes unexpected breakage: a. Wiederherstellen from checkpoint b. Investigate the hidden Abhaengigkeit c. Hinzufuegen it to die Schnittstelle archaeology (Step 3) d. Erstellen an explicit stub for the Abhaengigkeit e. Re-attempt dissolution
  4. Verfolgen dissolution progress:
    • Components remaining vs. dissolved
    • Imaginal discs extracted and verified portable
    • Unexpected Abhaengigkeiten discovered and handled

Erwartet: Systematic, verified dissolution of non-essential structure. After each layer, the remaining system is smaller, simpler, and still operational. Imaginal discs are preserved in portable form.

Bei Fehler: If dissolution causes cascading failures, the layer ordering is wrong — there are hidden Abhaengigkeiten deeper than expected. Stop, restore, remap Abhaengigkeiten, and re-sequence. If dissolution reveals that an "imaginal disc" is more complex than expected, allocate more extraction time for that capability.

Schritt 5: Vorbereiten the Foundation for Reconstruction

After dissolution, the remaining system sollte a minimal viable core plus extracted imaginal discs ready for reconstruction.

  1. Bewerten the post-dissolution state:
    • What remains? (minimal operational core + extracted capabilities)
    • Is the remaining system maintainable? (can das Team understand and modify it)
    • Are all imaginal discs accessible and verified? (portable, tested, documented)
  2. Erstellen the reconstruction manifest:
    • Auflisten each imaginal disc with its contract, data, and Testsuite
    • Angeben das Ziel architecture for reconstruction (or mark as "to be determined")
    • Identifizieren gaps: capabilities that were teilweise extracted or have quality concerns
  3. Handoff to reconstruction:
    • If das Ziel form is known: proceed to adapt-architecture with the minimal core as starting point
    • If das Ziel form is unknown: operate on the minimal core while das Ziel is designed
    • Either way: das System is now flexible enough to be reshaped

Erwartet: A minimal, maintainable system with clearly documented extracted capabilities. The foundation is clean and ready for reconstruction in whatever form is chosen.

Bei Fehler: If the post-dissolution system is less maintainable than expected, some essential structure was dissolved that should wurden preserved. Check the imaginal disc inventory — if a critical capability fehlt, it may still be recoverable from the archive. If the minimal core is too minimal to operate, some "replaceable tissue" was actually essential — restore it from the checkpoint.

Validierung

  • Imaginal discs are identified, extracted, and verified in portable form
  • Dissolution sequence is layered from outermost (no dependents) to core
  • Interface archaeology has captured actual (not just documented) behavior
  • Each dissolution layer has a verified checkpoint
  • No essential capability was lost waehrend dissolution
  • Post-dissolution system is minimal, maintainable, and operational
  • Reconstruction manifest documents extracted capabilities and gaps

Haeufige Stolperfallen

  • Dissolving ohne extracting: Removing a rigid component vor its essential capabilities are extracted destroys irreplaceable knowledge. Always extract imaginal discs first
  • Trusting documentation over observation: Documented interfaces often diverge from actual behavior. Interface archaeology (Step 3) reveals the truth; documentation shows the intent
  • Dissolving the core first: Removing load-bearing structures vor their dependents are dissolved causes cascading failure. Always work outside-in
  • Abschliessen dissolution: Dissolving everything to start from scratch sounds clean but loses institutional knowledge, battle-tested edge case handling, and operational continuity. Preserve imaginal discs
  • Dissolution as punishment: Dissolving a system "because it's bad" ohne a reconstruction plan creates a vacuum. Dissolution is the preparation for reconstruction, not an end in itself

Verwandte Skills

  • assess-form — prerequisite assessment that identifies rigidity and triggers dissolution
  • adapt-architecture — the reconstruction skill that follows dissolution
  • repair-damage — for systems that need targeted repair anstatt full dissolution
  • build-consensus — consensus vor major dissolution prevents team fragmentation
  • decommission-validated-system — formal decommissioning process for regulated systems
  • conduct-post-mortem — post-mortem analysis shares the investigative rigor of dissolution

GitHub 저장소

pjt222/agent-almanac
경로: i18n/de/skills/dissolve-form
0
agentsagentskillsai-assisted-developmentclaude-codeskillsteams

연관 스킬

llamaguard

기타

LlamaGuard는 폭력 및 혐오 발언 등 6가지 안전 범주에서 LLM 입력과 출력을 조정하기 위한 Meta의 70-80억 파라미터 모델입니다. 94-95% 정확도를 제공하며 vLLM, Hugging Face 또는 Amazon SageMaker를 사용해 배포할 수 있습니다. 이 기술을 사용하여 AI 애플리케이션에 콘텐츠 필터링 및 안전 가드레일을 손쉽게 통합하세요.

스킬 보기

cost-optimization

기타

이 Claude Skill은 리소스 적정화, 태깅 전략, 지출 분석을 통해 개발자들이 클라우드 비용을 최적화할 수 있도록 지원합니다. AWS, Azure, GCP에서 클라우드 비용을 절감하고 비용 거버넌스를 구현하기 위한 프레임워크를 제공합니다. 인프라 비용을 분석하거나, 리소스를 적정화하거나, 예산 제약을 충족해야 할 때 사용하세요.

스킬 보기

quantizing-models-bitsandbytes

기타

이 스킬은 bitsandbytes를 사용하여 LLM을 8비트 또는 4비트 정밀도로 양자화하며, 최소한의 정확도 손실로 50-75%의 메모리 감소를 달성합니다. 제한된 GPU 메모리에서 더 큰 모델을 실행하거나 추론을 가속화하는 데 이상적이며, INT8, NF4, FP4와 같은 형식을 지원합니다. 이 스킬은 HuggingFace Transformers와 통합되어 QLoRA 학습 및 8비트 옵티마이저를 가능하게 합니다.

스킬 보기

dispatching-parallel-agents

기타

이 Claude Skill은 3개 이상의 독립적인 문제를 동시에 조사하고 해결하기 위해 다중 에이전트를 배치합니다. 공유 상태나 의존성 없이 해결 가능한 무관련 장애 시나리오에 맞게 설계되었습니다. 핵심 기능은 병렬 문제 해결로, 각 독립 문제 영역마다 하나의 에이전트를 할당하여 효율성을 극대화합니다.

스킬 보기