1. Sample Context 사용량
❯ /context
⎿ Context Usage
⛁ ⛁ ⛁ ⛁ ⛁ ⛁ ⛁ ⛁ ⛁ ⛁ claude-opus-4-5-20251101 · 47k/200k tokens (24%)
⛁ ⛁ ⛁ ⛁ ⛁ ⛁ ⛁ ⛁ ⛁ ⛁
⛁ ⛁ ⛁ ⛁ ⛁ ⛶ ⛶ ⛶ ⛶ ⛶ Estimated usage by category
⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛁ System prompt: 2.3k tokens (1.2%)
⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛁ System tools: 19.9k tokens (9.9%)
⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛁ MCP tools: 8.8k tokens (4.4%)
⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛁ Custom agents: 4.5k tokens (2.3%)
⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛁ Skills: 8.9k tokens (4.5%)
⛶ ⛶ ⛶ ⛝ ⛝ ⛝ ⛝ ⛝ ⛝ ⛝ ⛁ Messages: 8.9k tokens (4.4%)
⛝ ⛝ ⛝ ⛝ ⛝ ⛝ ⛝ ⛝ ⛝ ⛝ ⛶ Free space: 114k (56.9%)
⛝ Autocompact buffer: 33.0k tokens (16.5%)
MCP tools · /mcp
└ mcp__ide__getDiagnostics: 111 tokens
└ mcp__ide__executeCode: 182 tokens
└ mcp__tavily__tavily-search: 625 tokens
└ mcp__tavily__tavily-extract: 207 tokens
└ mcp__plugin_context7_context7__resolve-library-id: 497 tokens
└ mcp__plugin_context7_context7__query-docs: 406 tokens
└ mcp__plugin_playwright_playwright__browser_close: 79 tokens
└ mcp__plugin_playwright_playwright__browser_resize: 128 tokens
└ mcp__plugin_playwright_playwright__browser_console_messages: 179 tokens
└ mcp__plugin_playwright_playwright__browser_handle_dialog: 132 tokens
└ mcp__plugin_playwright_playwright__browser_evaluate: 177 tokens
└ mcp__plugin_playwright_playwright__browser_file_upload: 132 tokens
└ mcp__plugin_playwright_playwright__browser_fill_form: 282 tokens
└ mcp__plugin_playwright_playwright__browser_install: 97 tokens
└ mcp__plugin_playwright_playwright__browser_press_key: 126 tokens
└ mcp__plugin_playwright_playwright__browser_type: 231 tokens
└ mcp__plugin_playwright_playwright__browser_navigate: 104 tokens
└ mcp__plugin_playwright_playwright__browser_navigate_back: 87 tokens
└ mcp__plugin_playwright_playwright__browser_network_requests: 168 tokens
└ mcp__plugin_playwright_playwright__browser_run_code: 170 tokens
└ mcp__plugin_playwright_playwright__browser_take_screenshot: 338 tokens
└ mcp__plugin_playwright_playwright__browser_snapshot: 115 tokens
└ mcp__plugin_playwright_playwright__browser_click: 256 tokens
└ mcp__plugin_playwright_playwright__browser_drag: 215 tokens
└ mcp__plugin_playwright_playwright__browser_hover: 139 tokens
└ mcp__plugin_playwright_playwright__browser_select_option: 187 tokens
└ mcp__plugin_playwright_playwright__browser_tabs: 158 tokens
└ mcp__plugin_playwright_playwright__browser_wait_for: 156 tokens
└ mcp__filesystem__read_file: 151 tokens
└ mcp__filesystem__read_text_file: 232 tokens
└ mcp__filesystem__read_media_file: 103 tokens
└ mcp__filesystem__read_multiple_files: 180 tokens
└ mcp__filesystem__write_file: 132 tokens
└ mcp__filesystem__edit_file: 229 tokens
└ mcp__filesystem__create_directory: 128 tokens
└ mcp__filesystem__list_directory: 130 tokens
└ mcp__filesystem__list_directory_with_sizes: 174 tokens
└ mcp__filesystem__directory_tree: 181 tokens
└ mcp__filesystem__move_file: 146 tokens
└ mcp__filesystem__search_files: 201 tokens
└ mcp__filesystem__get_file_info: 126 tokens
└ mcp__filesystem__list_allowed_directories: 107 tokens
└ mcp__context7__resolve-library-id: 492 tokens
└ mcp__context7__query-docs: 401 tokens
Custom agents · /agents
Plugin
└ bkit:pdca-iterator: 359 tokens
└ bkit:starter-guide: 310 tokens
└ bkit:pipeline-guide: 308 tokens
└ bkit:qa-monitor: 285 tokens
└ bkit:gap-detector: 277 tokens
└ bkit:code-analyzer: 272 tokens
└ bkit:report-generator: 268 tokens
└ bkit:enterprise-expert: 267 tokens
└ bkit:bkend-expert: 265 tokens
└ bkit:infra-architect: 257 tokens
└ superpowers:code-reviewer: 247 tokens
└ bkit:design-validator: 238 tokens
└ oh-my-claudecode:security-reviewer: 64 tokens
└ oh-my-claudecode:build-fixer: 63 tokens
└ oh-my-claudecode:tdd-guide: 61 tokens
└ oh-my-claudecode:code-reviewer: 54 tokens
└ oh-my-claudecode:build-fixer-low: 42 tokens
└ oh-my-claudecode:security-reviewer-low: 38 tokens
└ oh-my-claudecode:code-reviewer-low: 36 tokens
└ oh-my-claudecode:tdd-guide-low: 36 tokens
└ oh-my-claudecode:vision: 34 tokens
└ oh-my-claudecode:explore: 32 tokens
└ oh-my-claudecode:architect-medium: 31 tokens
└ oh-my-claudecode:architect: 31 tokens
└ oh-my-claudecode:explore-medium: 31 tokens
└ oh-my-claudecode:qa-tester-high: 31 tokens
└ oh-my-claudecode:writer: 31 tokens
└ oh-my-claudecode:designer: 30 tokens
└ oh-my-claudecode:qa-tester: 30 tokens
└ oh-my-claudecode:scientist-high: 30 tokens
└ oh-my-claudecode:architect-low: 29 tokens
└ oh-my-claudecode:designer-low: 29 tokens
└ oh-my-claudecode:executor-low: 28 tokens
└ oh-my-claudecode:explore-high: 28 tokens
└ oh-my-claudecode:scientist-low: 28 tokens
└ oh-my-claudecode:templates:base-agent: 28 tokens
└ oh-my-claudecode:templates:tier-instructions: 28 tokens
└ oh-my-claudecode:designer-high: 27 tokens
└ oh-my-claudecode:executor-high: 27 tokens
└ oh-my-claudecode:executor: 27 tokens
└ oh-my-claudecode:analyst: 26 tokens
└ oh-my-claudecode:planner: 26 tokens
└ oh-my-claudecode:researcher-low: 26 tokens
└ oh-my-claudecode:researcher: 26 tokens
└ oh-my-claudecode:scientist: 26 tokens
└ oh-my-claudecode:templates:README: 26 tokens
└ oh-my-claudecode:critic: 25 tokens
Skills · /skills
User
└ sc:recommend: 33 tokens
└ sc:pm: 31 tokens
└ sc:spec-panel: 31 tokens
└ sc:design: 26 tokens
└ sc:index: 26 tokens
└ sc:analyze: 26 tokens
└ sc:brainstorm: 25 tokens
└ sc:explain: 25 tokens
└ sc:implement: 25 tokens
└ sc:estimate: 25 tokens
└ sc:save: 24 tokens
└ sc:build: 24 tokens
└ sc:select-tool: 24 tokens
└ sc:sc: 23 tokens
└ sc:load: 23 tokens
└ sc:workflow: 23 tokens
└ sc:improve: 23 tokens
└ sc:troubleshoot: 23 tokens
└ sc:cleanup: 22 tokens
└ sc:document: 22 tokens
└ sc:task: 20 tokens
└ sc:git: 20 tokens
└ sc:reflect: 20 tokens
└ sc:spawn: 20 tokens
└ sc:test: 19 tokens
└ sc:research: 19 tokens
└ sc:business-panel: 17 tokens
└ sc:index-repo: 17 tokens
└ sc:help: 16 tokens
└ sc:README: 8 tokens
└ sc:agent: 6 tokens
Plugin
└ enterprise: 240 tokens
└ dynamic: 234 tokens
└ development-pipeline: 225 tokens
└ starter: 222 tokens
└ phase-8-review: 214 tokens
└ claude-code-learning: 206 tokens
└ data-context-extractor: 206 tokens
└ bkit-rules: 189 tokens
└ bkit-templates: 187 tokens
└ pdca: 181 tokens
└ code-review: 172 tokens
└ phase-5-design-system: 162 tokens
└ zero-script-qa: 162 tokens
└ phase-6-ui-integration: 159 tokens
└ phase-7-seo-security: 157 tokens
└ phase-2-convention: 147 tokens
└ mobile-app: 140 tokens
└ phase-1-schema: 140 tokens
└ phase-4-api: 134 tokens
└ desktop-app: 133 tokens
└ phase-9-deployment: 132 tokens
└ phase-3-mockup: 129 tokens
└ xlsx: 117 tokens
└ doc-coauthoring: 115 tokens
└ frontend-design: 108 tokens
└ docx: 96 tokens
└ internal-comms: 90 tokens
└ algorithmic-art: 89 tokens
└ web-artifacts-builder: 82 tokens
└ canvas-design: 80 tokens
└ mcp-builder: 76 tokens
└ sql-queries: 76 tokens
└ pptx: 73 tokens
└ theme-factory: 73 tokens
└ data-validation: 72 tokens
└ interactive-dashboard-builder: 72 tokens
└ data-exploration: 71 tokens
└ data-visualization: 70 tokens
└ statistical-analysis: 70 tokens
└ pdf: 68 tokens
└ brand-guidelines: 67 tokens
└ receiving-code-review: 67 tokens
└ verification-before-completion: 67 tokens
└ slack-gif-creator: 65 tokens
└ skill-creator: 64 tokens
└ finishing-a-development-branch: 61 tokens
└ bkit: 60 tokens
└ webapp-testing: 59 tokens
└ using-git-worktrees: 59 tokens
└ brainstorming: 56 tokens
└ using-superpowers: 47 tokens
└ dispatching-parallel-agents: 37 tokens
└ requesting-code-review: 36 tokens
└ Notion:create-page: 34 tokens
└ oh-my-claudecode:ultrawork: 33 tokens
└ ultrawork: 33 tokens
└ executing-plans: 33 tokens
└ oh-my-claudecode:cancel: 32 tokens
└ cancel: 32 tokens
└ subagent-driven-development: 31 tokens
└ systematic-debugging: 31 tokens
└ writing-skills: 31 tokens
└ writer-memory: 30 tokens
└ Notion:create-database-row: 29 tokens
└ test-driven-development: 29 tokens
└ oh-my-claudecode:ralph-init: 28 tokens
└ frontend-ui-ux: 28 tokens
└ git-master: 28 tokens
└ project-session-manager: 28 tokens
└ ralph-init: 28 tokens
└ writing-plans: 28 tokens
└ oh-my-claudecode:learn-about-omc: 26 tokens
└ oh-my-claudecode:research: 26 tokens
└ Notion:search: 26 tokens
└ local-skills-setup: 26 tokens
└ pipeline: 26 tokens
└ research: 26 tokens
└ oh-my-claudecode:omc-setup: 25 tokens
└ oh-my-claudecode:psm: 25 tokens
└ Notion:database-query: 25 tokens
└ learn-about-omc: 25 tokens
└ omc-setup: 25 tokens
└ oh-my-claudecode:deepinit: 24 tokens
└ oh-my-claudecode:ecomode: 24 tokens
└ oh-my-claudecode:ralph: 24 tokens
└ oh-my-claudecode:ralplan: 24 tokens
└ oh-my-claudecode:swarm: 24 tokens
└ oh-my-claudecode:ultrapilot: 24 tokens
└ Notion:create-task: 24 tokens
└ data:explore-data: 24 tokens
└ deepinit: 24 tokens
└ ecomode: 24 tokens
└ ralph: 24 tokens
└ ralplan: 24 tokens
└ swarm: 24 tokens
└ tdd: 23 tokens
└ oh-my-claudecode:hud: 22 tokens
└ oh-my-claudecode:mcp-setup: 22 tokens
└ oh-my-claudecode:ultraqa: 22 tokens
└ data:build-dashboard: 22 tokens
└ hud: 22 tokens
└ mcp-setup: 22 tokens
└ ultraqa: 22 tokens
└ oh-my-claudecode:pipeline: 21 tokens
└ data:validate: 21 tokens
└ oh-my-claudecode:autopilot: 20 tokens
└ oh-my-claudecode:build-fix: 20 tokens
└ oh-my-claudecode:learner: 20 tokens
└ autopilot: 20 tokens
└ build-fix: 20 tokens
└ learner: 20 tokens
└ ultrapilot: 20 tokens
└ oh-my-claudecode:doctor: 19 tokens
└ oh-my-claudecode:security-review: 19 tokens
└ doctor: 19 tokens
└ security-review: 19 tokens
└ skill: 19 tokens
└ oh-my-claudecode:note: 18 tokens
└ oh-my-claudecode:release: 18 tokens
└ Notion:find: 18 tokens
└ Notion:tasks:explain-diff: 18 tokens
└ data:analyze: 18 tokens
└ data:write-query: 18 tokens
└ note: 18 tokens
└ plan: 18 tokens
└ release: 18 tokens
└ data:create-viz: 17 tokens
└ orchestrate: 17 tokens
└ ralph-loop:help: 16 tokens
└ Notion:tasks:setup: 16 tokens
└ oh-my-claudecode:code-review: 15 tokens
└ oh-my-claudecode:help: 15 tokens
└ oh-my-claudecode:plan: 15 tokens
└ oh-my-claudecode:tdd: 15 tokens
└ code-review: 15 tokens
└ help: 15 tokens
└ ralph-loop:ralph-loop: 14 tokens
└ oh-my-claudecode:analyze: 14 tokens
└ Notion:tasks:build: 14 tokens
└ analyze: 14 tokens
└ oh-my-claudecode:deepsearch: 13 tokens
└ Notion:tasks:plan: 13 tokens
└ code-review:code-review: 13 tokens
└ deepsearch: 13 tokens
└ ralph-loop:cancel-ralph: 12 tokens
└ oh-my-claudecode:review: 12 tokens
└ review: 12 tokens 2. 각 항목 의미
⛁ System prompt: 2.3k tokens (1.2%)
- Claude Code의 고정값으로 변경 불가
- 어떻게 동작할지 정의
- 영향도 낮음
⛁ System tools: 19.9k tokens (9.9%)
- Claude Code 내장 도구들
- 코드 실행, 파일 읽기/쓰기
- 가장 많은 토큰 사용중
⛁ MCP tools: 8.8k tokens (4.4%)
- 외부 MCP 서버들
ide (실행/진단): 293 tokens
tavily (웹검색): 832 tokens
context7 (문서검색): 1,295 tokens
playwright (브라우저 자동화): 4,358 tokens
filesystem: 2,640 tokens⛁ Custom agents: 4.5k tokens (2.3%)
bkit 시리즈: 3,055 tokens
├─ pdca-iterator (359)
├─ starter-guide (310)
├─ pipeline-guide (308)
└─ 11개 더...
oh-my-claudecode 시리즈: 1,469 tokens
├─ 48개의 다양한 에이전트들
└─ 대부분 사용 안 함⛁ Skills: 8.9k tokens (4.5%)
User skills: 696 tokens (30개)
Plugin skills: 8,226 tokens (143개!) - 143개 Skills : 대부분 사용 안 함
- 실제 사용: 5% 미만 예상
- 8.9k 토큰 사용
3. Claude Code 리소스 최적화
실제 사용 여부 확인 방법
$ claude
❯ 지금까지 이 프로젝트에서 playwright MCP 도구를 사용한 적이 있어?
> 이 프로젝트에서 oh-my-claudecode 에이전트를 사용한 적 있어?MCP 정리
❯ /mcp
⎿ MCP dialog dismissed
❯ /context
⎿ Context Usage
⛁ ⛁ ⛁ ⛁ ⛁ ⛁ ⛁ ⛁ ⛁ ⛁ claude-sonnet-4-5-20250929 · 53k/200k tokens (27%)
⛁ ⛁ ⛁ ⛀ ⛁ ⛁ ⛁ ⛁ ⛁ ⛁
⛁ ⛁ ⛁ ⛁ ⛁ ⛁ ⛁ ⛁ ⛁ ⛁ Estimated usage by category
⛁ ⛁ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛁ System prompt: 2.3k tokens (1.2%)
⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛁ System tools: 19.9k tokens (9.9%)
⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛁ MCP tools: 5.1k tokens (2.6%)
⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛁ Custom agents: 4.5k tokens (2.3%)
⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛶ ⛁ Skills: 8.9k tokens (4.5%)
⛶ ⛶ ⛶ ⛝ ⛝ ⛝ ⛝ ⛝ ⛝ ⛝ ⛁ Messages: 24.5k tokens (12.3%)
⛝ ⛝ ⛝ ⛝ ⛝ ⛝ ⛝ ⛝ ⛝ ⛝ ⛶ Free space: 102k (50.8%)
⛝ Autocompact buffer: 33.0k tokens (16.5%)- mcp에서 필요없는 항목 disable
- MCP tools: 8.8k tokens -> 5.1k tokens 로 줄어듬.
Plugin에서 Agent, Skills 비활성화
❯ /plugins- 직접 들어가서 disable 적용
Data Engineer