godmode 플러그인
Engineering discipline for AI coding agents: question the request, verify against current docs instead of memory, plan, build test-first on the cheapest capable model, review with a second model family, prove it before saying done
godmode 플러그인 설치 방법
두 명령을 모두 Claude Code 안에서 실행하십시오. 첫 번째는 마켓플레이스를 한 번 추가하고, 두 번째는 플러그인을 설치합니다.
/plugin marketplace add NoobyGains/godmode
/plugin install godmode@godmodegodmode 구성 요소
플러그인 파일에서 읽어온 35개의 항목입니다.
Agents 1
@agent-godmode:name 으로 호출- @agent-godmode:code-reviewer
파일에 설명이 없습니다.
Skills 33
작업이 일치하면 Claude가 로드합니다- agent-briefing
Standardizes multi-agent communication through structured briefs, report files, and specific handoff protocols to prevent context loss.
- code-review
Enforces a two-stage review process to verify specification compliance and code quality before committing or merging changes.
- codebase-conventions
Ensures new code matches existing repository patterns by surveying precedents and shadowing their style, error handling, and architecture.
- completion-gate
Enforces a verification protocol that requires fresh evidence before an agent can claim work is finished, fixed, or deployed.
- cross-model-review
Enforces a second opinion from a different AI model family to catch blind spots before code is merged, released, or deployed.
- delegated-execution
Executes complex implementation plans by dispatching independent tasks to fresh subagent implementers and reviewers while maintaining a persistent ledger.
- doctor
Diagnoses the GodMode installation by running a diagnostic script to verify hooks, skill availability, and environment configuration.
- environment-awareness
Enforces a protocol for detecting OS, shell, package managers, and runtimes before executing shell commands to prevent environment mismatches.
- error-recovery
Provides a structured escalation framework for agents to break out of repetitive failure loops and handle debugging stagnation.
- fault-diagnosis
Provides a structured four-phase protocol to reproduce, isolate, and identify the root cause of software failures before implementing a fix.
- grounding
Enforces a protocol for verifying external facts like API interfaces, versions, and prices against primary sources before writing code or documentation.
- inline-execution
Enables an agent to execute a pre-defined implementation plan task-by-task using a test-first approach and a persistent ledger.
- intent-discovery
Converts user requests into approved technical designs and specifications through iterative questioning and structured documentation.
- knowledge-capture
Enables agents to save durable, non-obvious lessons, user preferences, and project conventions into native memory or documentation to persist across sessions.
- merge-protocol
Safely manages the final stage of a development branch by merging, opening pull requests, preserving, or abandoning work after verifying test suites.
- model-routing
Optimizes agentic workflows by assigning specific model tiers and effort levels to different roles to balance cost, performance, and token usage.
- parallel-execution
Enables an agent to solve multiple independent problems concurrently by dispatching focused agents to isolated worktrees.
- performance-tuning
Optimizes code and systems by using a measurement-driven protocol to identify and resolve latency, memory, and bundle size issues.
- project-bootstrap
Scaffolds a professional project foundation including version control, linting, formatting, testing, and CI workflows.
- quality-gates
Enforces project standards by managing linting, type checking, testing, coverage, build, size, dependency audits, complexity, and dead code checks.
- rationale
Evaluates new feature requests by analyzing problems, effort, alternatives, and hidden costs to provide informed decision options.
- reference-first
Ensures design decisions for APIs, schemas, and architectures are based on inspected, cited, and verified external or internal references.
- review-response
Evaluates and responds to technical code review feedback by verifying claims against the codebase before implementing any changes.
- security-protocol
Enforces secure coding practices by validating all external data, using safe APIs for sinks, managing secrets properly, and implementing robust authorization.
- system-design
Provides a structured protocol for selecting architecture, databases, APIs, and hosting based on current requirements and existing infrastructure.
- task-planning
Creates detailed implementation plans with atomic, verifiable tasks for multi-file features, refactors, or migrations.
- team-orchestration
Coordinates multiple autonomous agents in a single session to collaborate on complex tasks through mid-work messaging and shared task lists.
- test-first
Implements a strict red-green-refactor protocol to ensure all code changes are driven by failing tests before production code is written.
- ui-engineering
Builds and modifies frontend components, layouts, and forms using design systems with a focus on accessibility, state management, and responsiveness.
- ui-research
Researches layouts, design tokens, and component patterns to establish a visual foundation before any user interface code is written.
- using-godmode
Enforces engineering discipline by requiring agents to identify and invoke specific specialized skills before performing any action or response.
- workspace-isolation
Creates isolated git worktrees to protect the main checkout and enable parallel agent execution without file conflicts.
- writing-skills
Provides a standardized protocol for creating, testing, and validating agent skills using a baseline failure and verified pass methodology.
Hooks 1
다음 이벤트에서 실행됩니다- hooks
파일에 설명이 없습니다.
Claude Code에서 godmode 사용하기
이름으로 agent를 요청하세요:
@agent-godmode:code-reviewerSkills는 명령어가 필요하지 않습니다. 요청이 설명과 일치하면 Claude가 이를 읽습니다.
업데이트, 비활성화 또는 삭제
- 플러그인 업데이트 (shell)
- claude plugin update godmode@godmode
- marketplace 새로고침
- /plugin marketplace update godmode
- 일시적으로 비활성화
- /plugin disable godmode@godmode
- 삭제
- /plugin uninstall godmode@godmode
godmode의 스킬
- code-reviewNoobyGains/godmode
Enforces a two-stage review process to verify specification compliance and code quality before committing or merging changes.
2개 파일119스타 - cross-model-reviewNoobyGains/godmode
Enforces a second opinion from a different AI model family to catch blind spots before code is merged, released, or deployed.
2개 파일119스타 - delegated-executionNoobyGains/godmode
Executes complex implementation plans by dispatching independent tasks to fresh subagent implementers and reviewers while maintaining a persistent ledger.
5개 파일119스타 - environment-awarenessNoobyGains/godmode
Enforces a protocol for detecting OS, shell, package managers, and runtimes before executing shell commands to prevent environment mismatches.
2개 파일119스타 - fault-diagnosisNoobyGains/godmode
Provides a structured four-phase protocol to reproduce, isolate, and identify the root cause of software failures before implementing a fix.
5개 파일119스타 - intent-discoveryNoobyGains/godmode
Converts user requests into approved technical designs and specifications through iterative questioning and structured documentation.
2개 파일119스타 - project-bootstrapNoobyGains/godmode
Scaffolds a professional project foundation including version control, linting, formatting, testing, and CI workflows.
2개 파일119스타 - reference-firstNoobyGains/godmode
Ensures design decisions for APIs, schemas, and architectures are based on inspected, cited, and verified external or internal references.
5개 파일119스타 - security-protocolNoobyGains/godmode
Enforces secure coding practices by validating all external data, using safe APIs for sinks, managing secrets properly, and implementing robust authorization.
2개 파일119스타 - system-designNoobyGains/godmode
Provides a structured protocol for selecting architecture, databases, APIs, and hosting based on current requirements and existing infrastructure.
2개 파일119스타 - team-orchestrationNoobyGains/godmode
Coordinates multiple autonomous agents in a single session to collaborate on complex tasks through mid-work messaging and shared task lists.
2개 파일119스타 - test-firstNoobyGains/godmode
Implements a strict red-green-refactor protocol to ensure all code changes are driven by failing tests before production code is written.
2개 파일119스타 - ui-engineeringNoobyGains/godmode
Builds and modifies frontend components, layouts, and forms using design systems with a focus on accessibility, state management, and responsiveness.
3개 파일119스타 - ui-researchNoobyGains/godmode
Researches layouts, design tokens, and component patterns to establish a visual foundation before any user interface code is written.
4개 파일119스타 - using-godmodeNoobyGains/godmode
Enforces engineering discipline by requiring agents to identify and invoke specific specialized skills before performing any action or response.
7개 파일119스타 - writing-skillsNoobyGains/godmode
Provides a standardized protocol for creating, testing, and validating agent skills using a baseline failure and verified pass methodology.
2개 파일119스타 - agent-briefingNoobyGains/godmode
Standardizes multi-agent communication through structured briefs, report files, and specific handoff protocols to prevent context loss.
119스타 - codebase-conventionsNoobyGains/godmode
Ensures new code matches existing repository patterns by surveying precedents and shadowing their style, error handling, and architecture.
119스타 - completion-gateNoobyGains/godmode
Enforces a verification protocol that requires fresh evidence before an agent can claim work is finished, fixed, or deployed.
119스타 - doctorNoobyGains/godmode
Diagnoses the GodMode installation by running a diagnostic script to verify hooks, skill availability, and environment configuration.
119스타 - error-recoveryNoobyGains/godmode
Provides a structured escalation framework for agents to break out of repetitive failure loops and handle debugging stagnation.
119스타 - groundingNoobyGains/godmode
Enforces a protocol for verifying external facts like API interfaces, versions, and prices against primary sources before writing code or documentation.
119스타 - inline-executionNoobyGains/godmode
Enables an agent to execute a pre-defined implementation plan task-by-task using a test-first approach and a persistent ledger.
119스타 - knowledge-captureNoobyGains/godmode
Enables agents to save durable, non-obvious lessons, user preferences, and project conventions into native memory or documentation to persist across sessions.
119스타 - merge-protocolNoobyGains/godmode
Safely manages the final stage of a development branch by merging, opening pull requests, preserving, or abandoning work after verifying test suites.
119스타 - model-routingNoobyGains/godmode
Optimizes agentic workflows by assigning specific model tiers and effort levels to different roles to balance cost, performance, and token usage.
119스타 - parallel-executionNoobyGains/godmode
Enables an agent to solve multiple independent problems concurrently by dispatching focused agents to isolated worktrees.
119스타 - performance-tuningNoobyGains/godmode
Optimizes code and systems by using a measurement-driven protocol to identify and resolve latency, memory, and bundle size issues.
119스타 - quality-gatesNoobyGains/godmode
Enforces project standards by managing linting, type checking, testing, coverage, build, size, dependency audits, complexity, and dead code checks.
119스타 - rationaleNoobyGains/godmode
Evaluates new feature requests by analyzing problems, effort, alternatives, and hidden costs to provide informed decision options.
119스타
godmode 플러그인 질문
godmode 플러그인은 어떻게 설치하나요?
Claude Code에서 /plugin marketplace add NoobyGains/godmode를 실행한 다음 /plugin install godmode@godmode을 실행하십시오.
godmode 플러그인에는 무엇이 포함되나요?
godmode은(는) 1개 에이전트, 33개 스킬, 훅을(를) 묶습니다.
godmode에는 어떤 agent가 포함되어 있나요?
code-reviewer입니다. @agent-godmode:code-reviewer로 호출하세요.
godmode 플러그인을 어떻게 업데이트하거나 삭제하나요?
shell에서 claude plugin update godmode@godmode를 사용하여 업데이트하거나 /plugin의 Installed 탭에서 수행하세요. 삭제하려면 /plugin uninstall godmode@godmode을 사용하세요.