MCP HubMCP Hub
스킬 목록으로 돌아가기

adaptic

pjt222
업데이트됨 2 days ago
6 조회
17
2
17
GitHub에서 보기
메타aidesign

정보

Adpatic은 3개 이상의 영역에 걸쳐 파노라마적 종합을 달성하기 위해 5단계 개요 주기를 조율하는 마스터 스킬로, 순차적 타협이 아닌 통합적 이해를 생성합니다. 도메인 간 상호작용이 중요하거나 여러 이해관계자에게 영향을 미치는 주요 아키텍처 결정 전에 사용하세요. 이 스킬은 명상, 관찰, 게슈탈트 통합과 같은 단계를 복잡한 크로스도메인 문제에 대한 일관된 프로세스로 통합합니다.

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/adaptic

Claude Code에서 이 명령을 복사하여 붙여넣어 스킬을 설치하세요

문서

Adaptic

Compose the 5-step synoptic cycle to achieve panoramic synthesis across multiple domains. Where sequential analysis produces compromise ("a little of each"), the synoptic cycle produces integration — a unified understanding that holds all domains simultaneously and finds the emergent whole.

When to Use

  • A problem genuinely spans 3+ domains and the interactions between domains matter more than depth in any one
  • Sequential analysis (polymath style) has been tried but the synthesis feels like compromise rather than integration
  • Existing approaches feel like "a little of each" rather than a unified vision
  • Before major architectural decisions affecting multiple stakeholders
  • When domain experts disagree and the resolution lies between their perspectives, not within any one

When NOT to Use

  • Single-domain problems — use the domain agent directly
  • Well-understood trade-offs where polymath-style sequential analysis suffices
  • Self-care or wellness contexts — use the tending team instead
  • When speed matters more than depth — the full cycle requires sustained attention

Inputs

  • Required: The problem or question requiring multi-domain synthesis
  • Optional: Explicit list of domains to hold (default: auto-detect from problem context)
  • Optional: Depth setting — light, standard, or deep (default: standard)
  • Optional: Expression form — narrative, diagram, table, or recommendation (default: auto)

Configuration

settings:
  depth: standard          # light (skip meditate), standard, deep (extended perceive)
  domains: auto            # auto-detect or explicit list
  expression_form: auto    # narrative, diagram, table, recommendation

Procedure

Step 1: Clear — Empty the Workspace

Run the meditate skill to clear prior context, assumptions, and single-domain bias.

  1. Execute the full meditate procedure: prepare, anchor, observe distractions, close
  2. Pay special attention to domain bias — the tendency to frame the problem through whichever domain was most recently active
  3. Clear any premature solutions that arrived before the full picture was visible
  4. If depth: light is set, abbreviate to a brief context-clearing pause rather than the full meditation

Got: The workspace is empty. No domain has priority. No solution has been pre-selected. The agent is in a neutral, receptive state ready to hold multiple perspectives simultaneously.

If fail: If a particular domain keeps asserting itself as "the real problem," name that bias explicitly: "I notice I am framing this as primarily a [domain] problem." Naming the bias loosens its grip. If clearing fails entirely, the problem may genuinely be single-domain — reconsider whether the synoptic cycle is needed.

Step 2: Open — Enter Panoramic Mode

Run the expand-awareness skill to shift from narrow focus to wide-field perception.

  1. Inventory all domains relevant to the problem — do not pre-filter or rank them
  2. For each domain, note its core concerns, constraints, and values without evaluating
  3. Soften the focus: hold all domains in awareness simultaneously rather than cycling through them one at a time
  4. Resist the pull to "start solving" — this step is purely about opening the field of view
  5. If domains were provided explicitly in the inputs, use those as the starting set but remain open to discovering additional relevant domains

Got: A panoramic field is open. All relevant domains are held in awareness simultaneously. The agent can sense the full landscape without zooming into any single domain. The feeling is spacious rather than overwhelming.

If fail: If the domain list feels incomplete, ask: "What perspective is missing that would change the picture?" If simultaneous awareness collapses into sequential scanning (domain A, then B, then C), slow down — the goal is to hold the whole field, not to tour its parts. If more than 7 domains are active, group related domains into clusters to reduce cognitive load while maintaining breadth.

Step 3: Perceive — Notice Cross-Domain Patterns

While maintaining panoramic awareness, run observe and awareness to notice patterns, tensions, and resonances across all visible domains.

  1. Hold the panoramic field open from Step 2 — do not narrow focus
  2. Run observe to notice what is present: what patterns repeat across domains? what tensions exist between domains? what resonances connect seemingly unrelated concerns?
  3. Run awareness to notice what is not being seen: which domains are being subtly ignored? where are blind spots? what assumptions are operating below the surface?
  4. Record cross-domain observations without interpreting them yet:
    • Tensions: where domains pull in opposite directions
    • Resonances: where domains reinforce or echo each other
    • Gaps: where no domain addresses a concern that the whole picture reveals
    • Surprises: where a domain contributes something unexpected to the picture
  5. If depth: deep is set, extend this step — cycle through observe and awareness multiple times, allowing subtler patterns to surface

The critical discipline: perceive across all domains simultaneously, not each domain in turn. Sequential perception loses the cross-domain patterns that are the entire point of the synoptic cycle.

Got: A rich set of cross-domain observations — tensions, resonances, gaps, and surprises. These observations span the boundaries between domains rather than living within any single one. The agent has noticed something that would not be visible from any single domain's perspective.

If fail: If observations are all within single domains ("in domain A, I notice X"), the panoramic field has collapsed. Return to Step 2 and re-open. If no cross-domain patterns emerge, the problem may not require synoptic treatment — it may be genuinely decomposable into independent domain problems. If the perceive step produces an overwhelming number of observations, prioritize tensions (they are where integration happens).

Step 4: Integrate — Form the Emergent Whole

Run the integrate-gestalt skill to synthesize cross-domain observations into a unified understanding.

  1. Map the tensions identified in Step 3 — do not resolve them prematurely; hold them as creative constraints
  2. Find the figure: what unified understanding emerges when all observations are held together? This is not a compromise or average — it is a new pattern that includes but transcends the individual domain perspectives
  3. Test the whole: does the integrated understanding honor each domain's core concerns? Does it resolve tensions or merely paper over them?
  4. Name the insight in one clear statement — if it cannot be stated simply, the integration is not yet complete
  5. Verify that the insight is genuinely emergent: could it have been reached by analyzing domains sequentially? If yes, the synoptic cycle added no value and sequential analysis would have sufficed

Got: A single integrated understanding that holds all domains simultaneously. The insight feels like discovery rather than construction — it emerged from the whole rather than being assembled from parts. Each domain's core concerns are honored, and the tensions between domains are resolved rather than compromised.

If fail: If integration produces "a little of each domain" rather than a unified whole, the gestalt has not formed. Return to Step 3 and look for the tensions that are being avoided — integration happens through tension, not around it. If no gestalt forms after extended effort, decompose: find the 2-3 domains with the strongest tensions and integrate those first, then expand.

Step 5: Express — Communicate the Integrated Understanding

Run the express-insight skill to communicate the synthesis to the intended audience.

  1. Assess the audience: what domains are they familiar with? what framing will make the integrated insight accessible?
  2. Choose the expression form (or use the one specified in inputs):
    • Narrative: for audiences that need to understand the journey from parts to whole
    • Diagram: for audiences that need to see structural relationships
    • Table: for audiences that need to compare domain perspectives systematically
    • Recommendation: for audiences that need an actionable decision
  3. Express the integrated understanding with transparency: show which domains contributed, where tensions were resolved, and what the emergent insight adds beyond any single perspective
  4. Invite challenge: explicitly note which aspects of the integration are strongest and which are most speculative

Got: A clear, well-formed expression of the integrated understanding that is accessible to the intended audience. The expression shows its work — the audience can see how domain perspectives contributed to the whole. The form matches the audience's needs.

If fail: If the expression feels like a list of domain perspectives rather than an integrated whole, the insight from Step 4 has been lost in translation. Return to the one-statement summary from Step 4 and build the expression outward from that center. If the audience framing is wrong, ask: "Who needs this and what decision does it inform?"

Validation

  • Step 1 (Clear) was executed — prior context and domain bias were explicitly released
  • Step 2 (Open) produced a panoramic field holding 3+ domains simultaneously
  • Step 3 (Perceive) identified cross-domain patterns (not within-domain observations)
  • Step 4 (Integrate) produced a single emergent insight that transcends any individual domain
  • Step 5 (Express) communicated the insight in a form appropriate to the audience
  • The final output could not have been produced by sequential single-domain analysis
  • Each domain's core concerns are honored in the integrated understanding
  • Tensions between domains were resolved through integration, not compromise

Pitfalls

  • Sequential masquerading as simultaneous: Cycling through domains one at a time and then stapling the results together is not synoptic perception. The test: did the cross-domain interactions produce something new, or is the output a concatenation of domain analyses?
  • Premature integration: Jumping to a synthesis before the panoramic field has fully opened. Steps 2 and 3 build the perceptual foundation that makes genuine integration possible — rushing them produces shallow synthesis.
  • Compromise instead of emergence: Averaging domain perspectives ("50% security, 50% usability") is compromise, not integration. True integration finds a frame where both concerns are fully met, or it honestly names the irreducible trade-off.
  • Overuse on single-domain problems: Not every problem needs panoramic synthesis. If the problem lives cleanly in one domain, synoptic treatment adds overhead without value. The "When NOT to Use" criteria exist for a reason.
  • Losing the insight in expression: Step 4 produces a clear gestalt, but Step 5 fragments it back into a domain-by-domain list. Keep the integrated insight as the center of expression; domain details are supporting evidence, not the main structure.
  • Domain inflation: Artificially expanding the domain count to justify synoptic treatment. Three genuinely relevant domains produce better synthesis than seven domains where four are peripheral.

Related Skills

  • meditate — Step 1 of the cycle; clears context and establishes neutral starting state
  • expand-awareness — Step 2 of the cycle; shifts from narrow focus to panoramic perception
  • observe — used in Step 3; notices what is present across the field
  • awareness — used in Step 3; notices what is not being seen, reveals blind spots
  • integrate-gestalt — Step 4 of the cycle; forms the emergent whole from cross-domain patterns
  • express-insight — Step 5 of the cycle; communicates the integrated understanding

GitHub 저장소

pjt222/agent-almanac
경로: i18n/caveman-lite/skills/adaptic
0
agentsagentskillsai-assisted-developmentclaude-codeskillsteams

연관 스킬

content-collections

메타

이 스킬은 콘텐츠 콜렉션(Content Collections)을 위한 프로덕션 검증된 설정을 제공합니다. 콘텐츠 콜렉션은 Markdown/MDX 파일을 Zod 검증이 포함된 타입 안전한 데이터 콜렉션으로 변환해주는 TypeScript 최우선 도구입니다. 블로그, 문서 사이트 또는 콘텐츠 중심의 Vite + React 애플리케이션을 구축할 때 타입 안전성과 자동 콘텐츠 검증을 보장하기 위해 사용하세요. Vite 플러그인 구성과 MDX 컴파일부터 배포 최적화 및 스키마 검증에 이르기까지 모든 것을 다룹니다.

스킬 보기

polymarket

메타

이 스킬은 개발자들이 Polymarket 예측 시장 플랫폼을 활용한 애플리케이션을 구축할 수 있도록 지원하며, 거래 및 시장 데이터를 위한 API 통합 기능을 포함합니다. 또한 WebSocket을 통한 실시간 데이터 스트리밍을 제공하여 실시간 거래와 시장 활동을 모니터링할 수 있습니다. 이를 통해 거래 전략을 구현하거나 실시간 시장 업데이트를 처리하는 도구를 생성하는 데 활용할 수 있습니다.

스킬 보기

creating-opencode-plugins

메타

이 스킬은 개발자들이 명령어, 파일, LSP 작업 등 25개 이상의 이벤트 유형에 연결되는 OpenCode 플러그인을 만들 수 있도록 돕습니다. JavaScript/TypeScript 모듈을 위한 플러그인 구조, 이벤트 API 명세, 구현 패턴을 제공합니다. OpenCode AI 어시스턴트의 라이프사이클을 사용자 정의 이벤트 기반 로직으로 가로채거나, 모니터링하거나, 확장해야 할 때 사용하세요.

스킬 보기

sglang

메타

SGLang은 RadixAttention 프리픽스 캐싱을 활용하여 JSON, 정규식, 에이전트 워크플로우를 위한 고속 구조화 생성에 특화된 고성능 LLM 서빙 프레임워크입니다. 특히 반복되는 프리픽스가 있는 작업에서 상당히 빠른 추론 속도를 제공하여 복잡한 구조화 출력 및 다중 턴 대화에 이상적입니다. 제약 디코딩이 필요하거나 광범위한 프리픽스 공유가 있는 애플리케이션을 구축할 때는 vLLM과 같은 대안보다 SGLang을 선택하십시오.

스킬 보기