make-laten 1.0.0

package/README.md

@@ -0,0 +1,132 @@
+# make-laten
+
+Universal efficiency toolkit for AI coding agents — compress, cache, route, and optimize token usage.
+
+## What it does
+
+make-laten reduces token consumption when working with AI coding agents by compressing outputs, caching results, and routing operations intelligently.
+
+**Token savings:** 60-90% reduction across file reads, grep, git diff, and web fetches.
+
+## Install
+
+```bash
+npm install -g make-laten
+```
+
+## CLI Commands
+
+```bash
+# Compressed file read (89% savings)
+make-laten read src/index.ts
+
+# Compressed grep with file grouping
+make-laten grep "TODO" src/
+
+# Compressed git diff
+make-laten git diff
+make-laten git diff --staged
+make-laten git status
+
+# Web search (via DuckDuckGo)
+make-laten search "typescript generics"
+
+# Web fetch + semantic extraction + compression
+make-laten fetch https://docs.example.com
+
+# Cache management
+make-laten cache stats
+make-laten cache clear
+
+# Install adapters for detected agents
+make-laten install
+make-laten install --status
+make-laten install --uninstall
+```
+
+## Programmatic Usage
+
+```typescript
+import {
+  FileReadCompressor,
+  GrepCompressor,
+  GitDiffCompressor,
+  SemanticCache,
+  WebRouter,
+  getAdapter,
+  createInstaller
+} from 'make-laten'
+
+// Compress file reads
+const compressor = new FileReadCompressor()
+const result = await compressor.compress({
+  content: readFile('src/main.ts'),
+  filePath: 'src/main.ts',
+  language: 'typescript'
+})
+
+// Semantic web search
+const web = new WebRouter()
+const results = await web.search('how to use async await')
+const page = await web.fetch('https://example.com/docs')
+
+// Cache with cosine similarity
+const cache = new SemanticCache()
+cache.set('key1', { content: 'docs', embedding: [0.1, 0.2], metadata: {} })
+const similar = cache.search([0.1, 0.2], 0.8)
+
+// Agent adapters
+const installer = createInstaller()
+await installer.install()
+```
+
+## Modules
+
+| Module | Description |
+|--------|-------------|
+| `compress` | File read, grep, git diff, output compressors |
+| `cache` | L1 session cache, L2 cross-session, L3 semantic (cosine similarity) |
+| `graph` | Knowledge graph for operation patterns |
+| `route` | Smart routing between compression strategies |
+| `tool` | Tool router with pattern matching |
+| `adapter` | Agent adapters (Claude Code, Codex, Gemini, Cursor, etc.) |
+| `learn` | Pattern mining and failure learning |
+| `correct` | Auto-correction engine |
+| `web` | Web search, fetch, semantic extraction |
+| `middleware` | Middleware pipeline |
+| `plugin` | Plugin system with lifecycle |
+| `security` | Rate limiter |
+| `logging` | Structured logger |
+| `error` | Error handler with suggestions |
+| `metrics` | Counters, gauges, histograms |
+| `session` | Session manager with timeout |
+
+## Supported Agents
+
+make-laten works with 7+ AI coding agents:
+
+- **Claude Code** — hook-based interception
+- **Codex** — hook-based interception
+- **Gemini CLI** — hook-based interception
+- **Cursor** — rules injection
+- **Windsurf** — rules injection
+- **Cline** — rules injection
+- **GitHub Copilot** — rules injection
+
+```bash
+# Auto-detect and configure
+make-laten install
+```
+
+## Development
+
+```bash
+npm install
+npm test              # Run 201 tests
+npm run build         # Build CJS + ESM
+npm run typecheck     # Type check
+```
+
+## License
+
+MIT