{"items":[{"id":"cmugwi43h01sequ06w4x1h2q5","slug":"codeaashu-claude-code-claude-code-skill","name":"claude-code-skill","description":"Development conventions and architecture guide for the Claude Code CLI repository.","authorId":"gh:codeaashu","authorName":"codeaashu","version":"0.1.0","category":"Prompt","securityLevel":"Sandbox","downloadsCount":0,"githubStars":3354,"pricePerCall":0,"manifest":{"name":"claude-code-skill","tools":[],"category":"Prompt","entrypoint":{"type":"prompt"},"description":"Development conventions and architecture guide for the Claude Code CLI repository.","permissions":[],"systemPrompt":"# Claude Code — Repository Skill\n\n## Project Overview\n\nClaude Code is Anthropic's CLI tool for interacting with Claude from the terminal. It supports file editing, shell commands, git workflows, code review, multi-agent coordination, IDE integration (VS Code, JetBrains), and Model Context Protocol (MCP).\n\n**Codebase:** ~1,900 files, 512,000+ lines of TypeScript under `src/`.\n\n## Tech Stack\n\n| Component        | Technology                                      |\n|------------------|------------------------------------------------|\n| Language         | TypeScript (strict mode, ES modules)           |\n| Runtime          | Bun (JSX support, `bun:bundle` feature flags)  |\n| Terminal UI      | React + Ink (React for CLI)                    |\n| CLI Parser       | Commander.js (`@commander-js/extra-typings`)   |\n| API Client       | `@anthropic-ai/sdk`                            |\n| Validation       | Zod v4                                         |\n| Linter/Formatter | Biome                                          |\n| Analytics        | GrowthBook (feature flags & A/B testing)       |\n| Protocol         | Model Context Protocol (MCP)                   |\n\n## Architecture\n\n### Directory Map (`src/`)\n\n| Directory        | Purpose                                                         |\n|------------------|-----------------------------------------------------------------|\n| `commands/`      | ~50 slash commands (`/commit`, `/review`, `/config`, etc.)      |\n| `tools/`         | ~40 agent tools (Bash, FileRead, FileWrite, Glob, Grep, etc.)  |\n| `components/`    | ~140 Ink/React UI components for terminal rendering             |\n| `services/`      | External integrations (API, OAuth, MCP, LSP, analytics, plugins)|\n| `bridge/`        | Bidirectional IDE communication layer                           |\n| `state/`         | React context + custom store (AppState)                         |\n| `hooks/`         | React hooks (permissions, keybindings, commands, settings)      |\n| `types/`         | TypeScript type definitions                                     |\n| `utils/`         | Utilities (shell, file ops, permissions, config, git)           |\n| `screens/`       | Full-screen UIs (Doctor, REPL, Resume, Compact)                 |\n| `skills/`        | Bundled skills + skill loader system                            |\n| `plugins/`       | Plugin system (marketplace + bundled plugins)                   |\n| `coordinator/`   | Multi-agent coordination & supervisor logic                     |\n| `tasks/`         | Task management (shell tasks, agent tasks, teammates)           |\n| `context/`       | React context providers (notifications, stats, FPS)             |\n| `memdir/`        | Persistent memory system (CLAUDE.md, user/project memory)       |\n| `entrypoints/`   | Initialization logic, Agent SDK, MCP entry                      |\n| `voice/`         | Voice input/output (STT, keyterms)                              |\n| `vim/`           | Vim mode keybinding support                                     |\n| `schemas/`       | Zod configuration schemas                                       |\n| `keybindings/`   | Keybinding configuration & resolver                             |\n| `migrations/`    | Config migrations between versions                              |\n| `outputStyles/`  | Output formatting & theming                                     |\n| `query/`         | Query pipeline & processing                                     |\n| `server/`        | Server/daemon mode                                              |\n| `remote/`        | Remote session handling                                         |\n\n### Key Files\n\n| File                | Role                                                |\n|---------------------|-----------------------------------------------------|\n| `src/main.tsx`      | CLI entry point (Commander parser, startup profiling)|\n| `src/QueryEngine.ts`| Core LLM API caller (streaming, tool-call loops)    |\n| `src/Tool.ts`       | Tool type definitions & `buildTool` factory          |\n| `src/tools.ts`      | Tool registry & presets                              |\n| `src/commands.ts`   | Command registry                                     |\n| `src/context.ts`    | System/user context collection (git status, memory)  |\n| `src/cost-tracker.ts`| Token cost tracking                                 |\n\n### Entry Points & Initialization Sequence\n\n1. `src/main.tsx` — Commander CLI parser, startup profiling\n2. `src/entrypoints/init.ts` — Config, telemetry, OAuth, MDM\n3. `src/entrypoints/cli.tsx` — CLI session orchestration\n4. `src/entrypoints/mcp.ts` — MCP server mode\n5. `src/entrypoints/sdk/` — Agent SDK (programmatic API)\n6. `src/replLauncher.tsx` — REPL session launcher\n\nStartup performs parallel initialization: MDM policy reads, Keychain prefetch, feature flag checks, then core init.\n\n## Patterns & Conventions\n\n### Tool Definition\n\nEach tool lives in `src/tools/{ToolName}/` and uses `buildTool`:\n\n```typescript\nexport const MyTool = buildTool({\n  name: 'MyTool',\n  aliases: ['my_tool'],\n  description: 'What this tool does',\n  inputSchema: z.object({\n    param: z.string(),\n  }),\n  async call(args, context, canUseTool, parentMessage, onProgress) {\n    // Execute and return { data: result, newMessages?: [...] }\n  },\n  async checkPermissions(input, context) { /* Permission checks */ },\n  isConcurrencySafe(input) { /* Can run in parallel? */ },\n  isReadOnly(input) { /* Non-destructive? */ },\n  prompt(options) { /* System prompt injection */ },\n  renderToolUseMessage(input, options) { /* UI for invocation */ },\n  renderToolResultMessage(content, progressMessages, options) { /* UI for result */ },\n})\n```\n\n**Directory structure per tool:** `{ToolName}.ts` or `.tsx` (main), `UI.tsx` (rendering), `prompt.ts` (system prompt), plus utility files.\n\n### Command Definition\n\nCommands live in `src/commands/` and follow three types:\n\n- **PromptCommand** — Sends a formatted prompt with injected tools (most commands)\n- **LocalCommand** — Runs in-process, returns text\n- **LocalJSXCommand** — Runs in-process, returns React JSX\n\n```typescript\nconst command = {\n  type: 'prompt',\n  name: 'my-command',\n  description: 'What this command does',\n  progressMessage: 'working...',\n  allowedTools: ['Bash(git *)', 'FileRead(*)'],\n  source: 'builtin',\n  async getPromptForCommand(args, context) {\n    return [{ type: 'text', text: '...' }]\n  },\n} satisfies Command\n```\n\nCommands are registered in `src/commands.ts` and invoked via `/command-name` in the REPL.\n\n### Component Structure\n\n- Functional React components with Ink primitives (`Box`, `Text`, `useInput()`)\n- Styled with Chalk for terminal colors\n- React Compiler for optimized re-renders\n- Design system primitives in `src/components/design-system/`\n\n### State Management\n\n- `AppState` via React context + custom store (`src/state/AppStateStore.ts`)\n- Mutable state object passed to tool contexts\n- Selector functions for derived state\n- Change observers in `src/state/onChangeAppState.ts`\n\n### Permission System\n\n- **Modes:** `default` (prompt per operation), `plan` (show plan, ask once), `bypassPermissions` (auto-approve), `auto` (ML classifier)\n- **Rules:** Wildcard patterns — `Bash(git *)`, `FileEdit(/src/*)`\n- Tools implement `checkPermissions()` returning `{ granted: boolean, reason?, prompt? }`\n\n### Feature Flags & Build\n\nBun's `bun:bundle` feature flags enable dead-code elimination at build time:\n\n```typescript\nimport { feature } from 'bun:bundle'\nif (feature('PROACTIVE')) { /* proactive agent tools */ }\n```\n\nNotable flags: `PROACTIVE`, `KAIROS`, `BRIDGE_MODE`, `VOICE_MODE`, `COORDINATOR_MODE`, `DAEMON`, `WORKFLOW_SCRIPTS`.\n\nSome features are also gated via `process.env.USER_TYPE === 'ant'`.\n\n## Naming Conventions\n\n| Element      | Convention           | Example                          |\n|-------------|---------------------|----------------------------------|\n| Files       | PascalCase (exports) or kebab-case (commands) | `BashTool.tsx`, `commit-push-pr.ts` |\n| Components  | PascalCase           | `App.tsx`, `PromptInput.tsx`     |\n| Types       | PascalCase, suffix with Props/State/Context | `ToolUseContext`     |\n| Hooks       | `use` prefix         | `useCanUseTool`, `useSettings`   |\n| Constants   | SCREAMING_SNAKE_CASE | `MAX_TOKENS`, `DEFAULT_TIMEOUT_MS`|\n\n## Import Practices\n\n- ES modules with `.js` extensions (Bun convention)\n- Lazy imports for circular dependency breaking: `const getModule = () => require('./heavy.js')`\n- Conditional imports via feature flags or `process.env`\n- `biome-ignore` markers for manual import ordering where needed\n\n## Services\n\n| Service             | Path                          | Purpose                           |\n|--------------------|-------------------------------|-----------------------------------|\n| API                | `services/api/`               | Anthropic SDK client, file uploads|\n| MCP                | `services/mcp/`               | MCP client, tool/resource discovery|\n| OAuth              | `services/oauth/`             | OAuth 2.0 auth flow               |\n| LSP                | `services/lsp/`               | Language Server Protocol manager   |\n| Analytics          | `services/analytics/`         | GrowthBook, telemetry, events     |\n| Plugins            | `services/plugins/`           | Plugin loader, marketplace         |\n| Compact            | `services/compact/`           | Context compression                |\n| Policy Limits      | `services/policyLimits/`      | Org rate limits, quota checking    |\n| Remote Settings    | `services/remoteManagedSettings/` | Managed settings sync (Enterprise) |\n| Token Estimation   | `services/tokenEstimation.ts` | Token count estimation             |\n\n## Configuration\n\n**Settings locations:**\n- **Global:** `~/.claude/config.json`, `~/.claude/settings.json`\n- **Project:** `.claude/config.json`, `.claude/settings.json`\n- **System:** macOS Keychain + MDM, Windows Registry + MDM\n- **Managed:** Remote sync for Enterprise users\n\n## Guidelines\n\n1. Read relevant source files before making changes — understand existing patterns first.\n2. Follow the tool/command/component patterns above when adding new ones.\n3. Keep edits minimal and focused — avoid unnecessary refactoring.\n4. Use Zod for all input validation at system boundaries.\n5. Gate experimental features behind `bun:bundle` feature flags or env checks.\n6. Respect the permission system — tools that modify state must implement `checkPermissions()`.\n7. Use lazy imports when adding dependencies that could create circular references.\n8. Update this file as project conventions evolve.","schemaVersion":1},"repoUrl":"https://github.com/codeaashu/claude-code","tags":["aashuu","claude","claude-ai","claude-code","claude-code-leaked","claude-code-skill","claude-desktop","claude-leak","claude-skills","prompt"],"stats":{"installVelocity7d":0,"retentionRate":0,"executions":0,"rating":null},"origin":"github","source":{"repo":"claude-code","audit":{"files":["bun.lock","package-lock.json","package.json"],"binaries":[],"findings":[{"kind":"dependency","rule":"DP-01","message":"@opentelemetry/core@1.30.1 has a known vulnerability: OpenTelemetry Core: Unbounded memory allocation in W3C Baggage propagation.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-8988-4f7v-96qf · npm:@opentelemetry/core@1.30.1","severity":"medium"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability: axios Vulnerable to Full Man-in-the-Middle via Prototype Pollution Gadget in `config.proxy`.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-35jp-ww65-95wh · npm:axios@1.14.0","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability: axios Vulnerable to Credential Theft and Response Hijacking via Prototype Pollution Gadget in Config Merge.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-3g43-6gmg-66jw · npm:axios@1.14.0","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability: Axios has a NO_PROXY Hostname Normalization Bypass that Leads to SSRF.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-3p68-rc4w-qgx5 · npm:axios@1.14.0","severity":"medium"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability: Axios: Invisible JSON Response Tampering via Prototype Pollution Gadget in `parseReviver`.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-3w6x-2g7m-8v23 · npm:axios@1.14.0","severity":"medium"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability: Axios: Excessive recursion in formDataToJSON can cause denial of service.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-42h9-826w-cgv3 · npm:axios@1.14.0","severity":"medium"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability: Axios: CRLF Injection in multipart/form-data body via unsanitized blob.type in formDataToStream.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-445q-vr5w-6q77 · npm:axios@1.14.0","severity":"medium"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability: Axios' HTTP adapter-streamed uploads bypass maxBodyLength when maxRedirects: 0.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-5c9x-8gcm-mpgx · npm:axios@1.14.0","severity":"medium"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability: Axios: unbounded recursion in toFormData causes DoS via deeply nested request data.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-62hf-57xw-28j9 · npm:axios@1.14.0","severity":"medium"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability: Axios: Header Injection via Prototype Pollution.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-6chq-wfr3-2hj9 · npm:axios@1.14.0","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-777c-7fjr-54vf · npm:axios@1.14.0","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-7q8q-rj6j-mhjq · npm:axios@1.14.0","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-898c-q2cr-xwhg · npm:axios@1.14.0","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-fvcv-3m26-pcqx · npm:axios@1.14.0","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-hfxv-24rg-xrqf · npm:axios@1.14.0","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-j5f8-grm9-p9fc · npm:axios@1.14.0","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-jqh4-m9w3-8hp9 · npm:axios@1.14.0","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-m7pr-hjqh-92cm · npm:axios@1.14.0","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-mmx7-hfxf-jppx · npm:axios@1.14.0","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-mwf2-3pr3-8698 · npm:axios@1.14.0","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-p92q-9vqr-4j8v · npm:axios@1.14.0","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-pf86-5x62-jrwf · npm:axios@1.14.0","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-pmv8-rq9r-6j72 · npm:axios@1.14.0","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-pmwg-cvhr-8vh7 · npm:axios@1.14.0","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-q8qp-cvcw-x6jj · npm:axios@1.14.0","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-vf2m-468p-8v99 · npm:axios@1.14.0","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-w9j2-pvgh-6h63 · npm:axios@1.14.0","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-xhjh-pmcv-23jw · npm:axios@1.14.0","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"axios@1.14.0 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-xx6v-rp6x-q39c · npm:axios@1.14.0","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"diff@7.0.0 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-73rr-hh4g-fpgx · npm:diff@7.0.0","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"lodash-es@4.17.23 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-f23m-r3pf-42rh · npm:lodash-es@4.17.23","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"lodash-es@4.17.23 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-r5fr-rjxr-66jc · npm:lodash-es@4.17.23","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"undici@7.24.6 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-35p6-xmwp-9g52 · npm:undici@7.24.6","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"undici@7.24.6 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-4cwx-7wf7-3272 · npm:undici@7.24.6","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"undici@7.24.6 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-8xcm-r25x-g524 · npm:undici@7.24.6","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"undici@7.24.6 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-g8m3-5g58-fq7m · npm:undici@7.24.6","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"undici@7.24.6 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-hm92-r4w5-c3mj · npm:undici@7.24.6","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"undici@7.24.6 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-jr45-8vmc-qm54 · npm:undici@7.24.6","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"undici@7.24.6 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-m8rv-5g2x-5cg5 · npm:undici@7.24.6","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"undici@7.24.6 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-p88m-4jfj-68fv · npm:undici@7.24.6","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"undici@7.24.6 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-pr7r-676h-xcf6 · npm:undici@7.24.6","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"undici@7.24.6 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-v3r7-h72x-cjcm · npm:undici@7.24.6","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"undici@7.24.6 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-vmh5-mc38-953g · npm:undici@7.24.6","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"undici@7.24.6 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-vxpw-j846-p89q · npm:undici@7.24.6","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"ws@8.20.0 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-58qx-3vcg-4xpx · npm:ws@8.20.0","severity":"high"},{"kind":"dependency","rule":"DP-01","message":"ws@8.20.0 has a known vulnerability.","surface":"bun.lock, package-lock.json, package.json","evidence":"GHSA-96hv-2xvq-fx4p · npm:ws@8.20.0","severity":"high"}],"packages":49,"auditedAt":"2026-09-25T11:52:11.817Z","lockfiles":["bun.lock","package-lock.json"]},"forks":3674,"owner":"codeaashu","stars":3354,"topics":["aashuu","claude","claude-ai","claude-code","claude-code-leaked","claude-code-skill","claude-desktop","claude-leak","claude-skills"],"license":null,"fullName":"codeaashu/claude-code","homepage":null,"language":"TypeScript","pushedAt":"2026-08-29T20:16:27Z","avatarUrl":"https://avatars.githubusercontent.com/u/130897584?v=4","crawledAt":"2026-09-25T11:51:57.794Z","openIssues":16,"manifestFile":"SKILL.md","manifestPath":"Skill.md","defaultBranch":"main"},"readme":"# Claude Code — Repository Skill\n\n## Project Overview\n\nClaude Code is Anthropic's CLI tool for interacting with Claude from the terminal. It supports file editing, shell commands, git workflows, code review, multi-agent coordination, IDE integration (VS Code, JetBrains), and Model Context Protocol (MCP).\n\n**Codebase:** ~1,900 files, 512,000+ lines of TypeScript under `src/`.\n\n## Tech Stack\n\n| Component        | Technology                                      |\n|------------------|------------------------------------------------|\n| Language         | TypeScript (strict mode, ES modules)           |\n| Runtime          | Bun (JSX support, `bun:bundle` feature flags)  |\n| Terminal UI      | React + Ink (React for CLI)                    |\n| CLI Parser       | Commander.js (`@commander-js/extra-typings`)   |\n| API Client       | `@anthropic-ai/sdk`                            |\n| Validation       | Zod v4                                         |\n| Linter/Formatter | Biome                                          |\n| Analytics        | GrowthBook (feature flags & A/B testing)       |\n| Protocol         | Model Context Protocol (MCP)                   |\n\n## Architecture\n\n### Directory Map (`src/`)\n\n| Directory        | Purpose                                                         |\n|------------------|-----------------------------------------------------------------|\n| `commands/`      | ~50 slash commands (`/commit`, `/review`, `/config`, etc.)      |\n| `tools/`         | ~40 agent tools (Bash, FileRead, FileWrite, Glob, Grep, etc.)  |\n| `components/`    | ~140 Ink/React UI components for terminal rendering             |\n| `services/`      | External integrations (API, OAuth, MCP, LSP, analytics, plugins)|\n| `bridge/`        | Bidirectional IDE communication layer                           |\n| `state/`         | React context + custom store (AppState)                         |\n| `hooks/`         | React hooks (permissions, keybindings, commands, settings)      |\n| `types/`         | TypeScript type definitions                                     |\n| `utils/`         | Utilities (shell, file ops, permissions, config, git)           |\n| `screens/`       | Full-screen UIs (Doctor, REPL, Resume, Compact)                 |\n| `skills/`        | Bundled skills + skill loader system                            |\n| `plugins/`       | Plugin system (marketplace + bundled plugins)                   |\n| `coordinator/`   | Multi-agent coordination & supervisor logic                     |\n| `tasks/`         | Task management (shell tasks, agent tasks, teammates)           |\n| `context/`       | React context providers (notifications, stats, FPS)             |\n| `memdir/`        | Persistent memory system (CLAUDE.md, user/project memory)       |\n| `entrypoints/`   | Initialization logic, Agent SDK, MCP entry                      |\n| `voice/`         | Voice input/output (STT, keyterms)                              |\n| `vim/`           | Vim mode keybinding support                                     |\n| `schemas/`       | Zod configuration schemas                                       |\n| `keybindings/`   | Keybinding configuration & resolver                             |\n| `migrations/`    | Config migrations between versions                              |\n| `outputStyles/`  | Output formatting & theming                                     |\n| `query/`         | Query pipeline & processing                                     |\n| `server/`        | Server/daemon mode                                              |\n| `remote/`        | Remote session handling                                         |\n\n### Key Files\n\n| File                | Role                                                |\n|---------------------|-----------------------------------------------------|\n| `src/main.tsx`      | CLI entry point (Commander parser, startup profiling)|\n| `src/QueryEngine.ts`| Core LLM API caller (streaming, tool-call loops)    |\n| `src/Tool.ts`       | T","createdAt":"2026-09-25T11:52:11.837Z","updatedAt":"2026-09-25T11:52:11.837Z"}],"total":1,"limit":24,"offset":0}