lynkr 9.7.0 → 9.7.1

package/README.md

@@ -1,6 +1,6 @@
 # Lynkr
 
-### The AI coding proxy that compresses tokens before they hit the model.
+### An LLM Gateway which optimises your token usage.
 
 **87.6% fewer tokens on JSON tool results. 53% fewer tokens on tool-heavy requests. 171ms semantic cache hits. Zero code changes.**
 
@@ -31,20 +31,11 @@
 ```bash
 npm install -g lynkr
 
+
 # Claude Code Pro/Max
 lynkr wrap claude
 
-# GitHub Copilot
-lynkr wrap copilot
-
-# Aider
-lynkr wrap aider
 
-# Cursor
-lynkr wrap cursor
-
-# OpenAI Codex
-lynkr wrap codex
 ```
 
 **Wrapping gives you:**
@@ -68,96 +59,31 @@ npm install -g lynkr
 
 ### 2. Configure Lynkr
 
-First run creates a `.env` file. Edit it with your provider settings.
-
-**Option A: Free & Local (Ollama) - Recommended for Testing**
+The fastest path is the interactive wizard:
 
 ```bash
-# Install Ollama first: https://ollama.com
-ollama pull qwen2.5-coder:latest
+lynkr init
 ```
 
-Create/edit `.env` in your project directory:
-```bash
-# Provider
-MODEL_PROVIDER=ollama
-FALLBACK_ENABLED=false
-
-# Ollama Configuration
-OLLAMA_ENDPOINT=http://localhost:11434
-OLLAMA_MODEL=qwen2.5-coder:latest
-
-# Server
-PORT=8081
-
-# Optional: Limits (remove for unlimited)
-POLICY_MAX_STEPS=50
-POLICY_MAX_TOOL_CALLS=100
-
-# Disable overly strict command filtering
-POLICY_SAFE_COMMANDS_ENABLED=false
-```
+It asks four questions — usage mode (Claude Pro/Max via wrap, or direct API keys), tier picks for SIMPLE/MEDIUM/COMPLEX/REASONING across the 12 supported providers, credentials for what you chose, and a few routing-intelligence knobs — then writes a fully-populated `.env` with sensible production defaults for everything else (caching, compression, policy budgets, MCP sandbox, agents, rate limiting).
 
-**Option B: Cloud (OpenRouter) - Recommended for Production**
+Useful flags:
 
 ```bash
-# Get API key from https://openrouter.ai
+lynkr init --force                # overwrite an existing .env
 ```
 
-Create/edit `.env`:
-```bash
-# Provider
-MODEL_PROVIDER=openrouter
-OPENROUTER_API_KEY=sk-or-v1-your-key-here
-FALLBACK_ENABLED=false
-
-# Server
-PORT=8081
+See [`docs/init.md`](docs/init.md) for the full wizard reference.
 
-# Optional: Limits (remove for unlimited)
-POLICY_MAX_STEPS=50
-POLICY_MAX_TOOL_CALLS=100
+If you'd rather configure by hand, the manual options below still work — copy `.env.example` to `.env` and edit it directly:
 
-# Optional: Enable caching
-PROMPT_CACHE_ENABLED=true
-SEMANTIC_CACHE_ENABLED=true
-```
-
-**Option C: Enterprise (AWS Bedrock)**
+**Option A: Free & Local (Ollama) - Recommended for Testing**
 
-Create/edit `.env`:
 ```bash
-# Provider
-MODEL_PROVIDER=bedrock
-AWS_BEDROCK_API_KEY=your-aws-key
-AWS_BEDROCK_MODEL_ID=anthropic.claude-3-5-sonnet-20241022-v2:0
-FALLBACK_ENABLED=false
-
-# Server
-PORT=8081
-
-# Optional: Limits (remove for unlimited)
-POLICY_MAX_STEPS=50
-POLICY_MAX_TOOL_CALLS=100
+# Install Ollama first: https://ollama.com
+ollama pull qwen2.5-coder:latest
 ```
 
-**Option D: Enterprise (Databricks)**
-
-Create/edit `.env`:
-```bash
-# Provider
-MODEL_PROVIDER=databricks
-DATABRICKS_API_BASE=https://your-workspace.cloud.databricks.com
-DATABRICKS_API_KEY=your-token
-FALLBACK_ENABLED=false
-
-# Server
-PORT=8081
-
-# Optional: Limits (remove for unlimited)
-POLICY_MAX_STEPS=50
-POLICY_MAX_TOOL_CALLS=100
-```
 
 Then start Lynkr:
 
@@ -167,21 +93,6 @@ lynkr start
 
 ### 3. Connect Your Tool
 
-**Claude Code**
-
-**Windows (Command Prompt):**
-```cmd
-set ANTHROPIC_BASE_URL=http://localhost:8081
-set ANTHROPIC_API_KEY=dummy
-claude "write a hello world in python"
-```
-
-**Linux/macOS:**
-```bash
-export ANTHROPIC_BASE_URL=http://localhost:8081
-export ANTHROPIC_API_KEY=dummy
-claude "write a hello world in python"
-```
 
 **Cursor IDE**
 - Settings → Models → Override Base URL
@@ -338,223 +249,6 @@ Tier configuration is strictly authoritative — bandit exploration is constrain
 
 ---
 
-## Complete .env Examples
-
-### MVP: Minimal Working Setup (Ollama)
-
-Copy-paste ready configuration for immediate use:
-
-```bash
-# .env - Minimal Ollama Setup
-
-# ============================================
-# REQUIRED: Provider Configuration
-# ============================================
-MODEL_PROVIDER=ollama
-FALLBACK_ENABLED=false
-
-# ============================================
-# REQUIRED: Ollama Settings
-# ============================================
-OLLAMA_ENDPOINT=http://localhost:11434
-OLLAMA_MODEL=qwen2.5-coder:latest
-
-# ============================================
-# REQUIRED: Server Configuration
-# ============================================
-PORT=8081
-HOST=0.0.0.0
-
-# ============================================
-# REQUIRED: Claude Code/Cursor Compatibility
-# ============================================
-POLICY_MAX_STEPS=50
-POLICY_MAX_TOOL_CALLS=100
-POLICY_SAFE_COMMANDS_ENABLED=false
-
-# ============================================
-# OPTIONAL: Performance (Recommended)
-# ============================================
-LOG_LEVEL=warn
-LOAD_SHEDDING_ENABLED=true
-LOAD_SHEDDING_HEAP_THRESHOLD=0.85
-```
-
-**Steps:**
-1. Install Ollama: `curl -fsSL https://ollama.com/install.sh | sh`
-2. Pull model: `ollama pull qwen2.5-coder:latest`
-3. Copy above to `.env` in your project directory
-4. Run: `lynkr start`
-
----
-
-### Production: Cloud with Tier Routing (OpenRouter)
-
-Optimized for cost savings with smart routing:
-
-```bash
-# .env - Production OpenRouter Setup
-
-# ============================================
-# REQUIRED: Provider Configuration
-# ============================================
-MODEL_PROVIDER=openrouter
-OPENROUTER_API_KEY=sk-or-v1-your-key-here
-FALLBACK_ENABLED=false
-
-# ============================================
-# REQUIRED: Server Configuration
-# ============================================
-PORT=8081
-HOST=0.0.0.0
-
-# ============================================
-# TIER ROUTING: Smart Cost Optimization
-# ============================================
-# Simple queries → Cheap/fast model
-TIER_SIMPLE=openrouter:google/gemini-flash-1.5
-
-# Normal coding → Balanced model
-TIER_MEDIUM=openrouter:anthropic/claude-3.5-sonnet
-
-# Complex refactoring → Powerful model
-TIER_COMPLEX=openrouter:anthropic/claude-opus-4
-
-# Deep reasoning → Most capable model
-TIER_REASONING=openrouter:anthropic/claude-opus-4
-
-# ============================================
-# REQUIRED: Claude Code/Cursor Compatibility
-# ============================================
-POLICY_MAX_STEPS=50
-POLICY_MAX_TOOL_CALLS=100
-POLICY_SAFE_COMMANDS_ENABLED=false
-
-# ============================================
-# OPTIONAL: Token Optimization (60-80% savings)
-# ============================================
-PROMPT_CACHE_ENABLED=true
-SEMANTIC_CACHE_ENABLED=true
-SEMANTIC_CACHE_THRESHOLD=0.95
-TOOL_INJECTION_ENABLED=false
-
-# ============================================
-# OPTIONAL: Performance Tuning
-# ============================================
-LOG_LEVEL=warn
-LOAD_SHEDDING_ENABLED=true
-LOAD_SHEDDING_HEAP_THRESHOLD=0.85
-```
-
-**Expected savings:** 70-90% of requests use Gemini Flash ($). Only 10-30% use Claude Opus ($$$).
-
----
-
-### Enterprise: Databricks Foundation Models
-
-For teams using Databricks Model Serving:
-
-```bash
-# .env - Enterprise Databricks Setup
-
-# ============================================
-# REQUIRED: Provider Configuration
-# ============================================
-MODEL_PROVIDER=databricks
-DATABRICKS_API_BASE=https://your-workspace.cloud.databricks.com
-DATABRICKS_API_KEY=dapi1234567890abcdef
-FALLBACK_ENABLED=false
-
-# ============================================
-# REQUIRED: Model Configuration
-# ============================================
-# Option 1: Single model (no tier routing)
-DATABRICKS_MODEL=databricks-meta-llama-3-1-405b-instruct
-
-# Option 2: Tier routing (comment out above, uncomment below)
-# TIER_SIMPLE=databricks:databricks-meta-llama-3-1-70b-instruct
-# TIER_MEDIUM=databricks:databricks-claude-sonnet-4-5
-# TIER_COMPLEX=databricks:databricks-claude-opus-4-6
-# TIER_REASONING=databricks:databricks-claude-opus-4-6
-
-# ============================================
-# REQUIRED: Server Configuration
-# ============================================
-PORT=8081
-HOST=0.0.0.0
-
-# ============================================
-# REQUIRED: Claude Code/Cursor Compatibility
-# ============================================
-POLICY_MAX_STEPS=50
-POLICY_MAX_TOOL_CALLS=100
-POLICY_SAFE_COMMANDS_ENABLED=false
-
-# ============================================
-# OPTIONAL: Enterprise Features
-# ============================================
-LOG_LEVEL=info
-LOAD_SHEDDING_ENABLED=true
-LOAD_SHEDDING_HEAP_THRESHOLD=0.85
-
-# Optional: Metrics for monitoring
-# PROMETHEUS_METRICS_ENABLED=true
-```
-
----
-
-### Hybrid: Local + Cloud Fallback
-
-Use free Ollama, fallback to cloud when needed:
-
-```bash
-# .env - Hybrid Setup (Advanced)
-
-# ============================================
-# PRIMARY: Local Ollama
-# ============================================
-MODEL_PROVIDER=ollama
-OLLAMA_ENDPOINT=http://localhost:11434
-OLLAMA_MODEL=qwen2.5-coder:latest
-
-# ============================================
-# FALLBACK: Cloud Provider
-# ============================================
-FALLBACK_ENABLED=true
-FALLBACK_PROVIDER=openrouter
-OPENROUTER_API_KEY=sk-or-v1-your-key-here
-
-# ============================================
-# TIER ROUTING: Mix Local + Cloud
-# ============================================
-TIER_SIMPLE=ollama:qwen2.5:3b
-TIER_MEDIUM=ollama:qwen2.5:7b
-TIER_COMPLEX=openrouter:anthropic/claude-3.5-sonnet
-TIER_REASONING=openrouter:anthropic/claude-opus-4
-
-# ============================================
-# REQUIRED: Server Configuration
-# ============================================
-PORT=8081
-HOST=0.0.0.0
-
-# ============================================
-# REQUIRED: Claude Code/Cursor Compatibility
-# ============================================
-POLICY_MAX_STEPS=50
-POLICY_MAX_TOOL_CALLS=100
-POLICY_SAFE_COMMANDS_ENABLED=false
-
-# ============================================
-# OPTIONAL: Performance
-# ============================================
-LOG_LEVEL=warn
-LOAD_SHEDDING_ENABLED=true
-```
-
-**Best of both worlds:** 80% of requests stay local (free). Complex tasks use cloud (paid).
-
----
 
 ## Common Issues & Fixes
 

package/bin/cli.js

@@ -9,11 +9,15 @@ const SUBCOMMANDS = {
   usage:      path.join(__dirname, "lynkr-usage.js"),
   trajectory: path.join(__dirname, "lynkr-trajectory.js"),
   wrap:       path.join(__dirname, "wrap.js"),
+  init:       path.join(__dirname, "lynkr-init.js"),
 };
 
 const sub = process.argv[2];
 if (sub && Object.prototype.hasOwnProperty.call(SUBCOMMANDS, sub)) {
   process.argv.splice(2, 1); // drop the subcommand token so the script's own arg parser is happy
+  // Subcommand scripts check this to decide whether to invoke their main()
+  // when they're require()'d (vs being loaded by a test for unit-checking).
+  process.env._LYNKR_SUBCMD = sub;
   require(SUBCOMMANDS[sub]);
   return;
 }

package/bin/lynkr-init.js

@@ -0,0 +1,679 @@
+#!/usr/bin/env node
+/**
+ * `lynkr init` — interactive setup wizard that produces a working .env.
+ *
+ * Walks the user through:
+ *   1. Usage mode (Claude Pro/Max subscription via wrap, or API-key direct).
+ *   2. Per-tier model selection across all supported providers.
+ *   3. Routing-intelligence knobs (visible badge, intent window, decay).
+ *   4. Credential collection (re-uses values across tiers, never asks twice).
+ *
+ * Usage:
+ *   lynkr init                        # interactive
+ *   lynkr init --force                # overwrite existing .env
+ *   lynkr init --output=<path>        # write to <path> instead of .env
+ *   lynkr init --dry-run              # print to stdout, don't write
+ *   lynkr init --help
+ *
+ * @module bin/lynkr-init
+ */
+
+const fs = require('fs');
+const path = require('path');
+const readline = require('readline');
+
+// ──────────────────────────────────────────────────────────────────────────────
+// Provider schema
+// ──────────────────────────────────────────────────────────────────────────────
+//
+// Canonical list pulled from src/config/index.js SUPPORTED_MODEL_PROVIDERS.
+// Each entry lists the credential env vars the user needs to supply and any
+// model/endpoint extras with sensible defaults. Local providers (no creds) are
+// flagged so the wizard skips the credential prompt.
+
+const PROVIDERS = {
+  ollama: {
+    label: 'Ollama (local, free)',
+    local: true,
+    creds: [],
+    extras: [
+      { key: 'OLLAMA_ENDPOINT', label: 'endpoint', default: 'http://localhost:11434' },
+    ],
+    defaultModel: 'qwen2.5-coder:latest',
+  },
+  llamacpp: {
+    label: 'llama.cpp (local)',
+    local: true,
+    creds: [],
+    extras: [
+      { key: 'LLAMACPP_ENDPOINT', label: 'endpoint', default: 'http://localhost:8080' },
+    ],
+    defaultModel: 'qwen2.5-coder',
+  },
+  lmstudio: {
+    label: 'LM Studio (local)',
+    local: true,
+    creds: [],
+    extras: [
+      { key: 'LMSTUDIO_ENDPOINT', label: 'endpoint', default: 'http://localhost:1234/v1' },
+    ],
+    defaultModel: 'qwen2.5-coder',
+  },
+  'azure-anthropic': {
+    label: 'Azure Anthropic (Claude via Azure)',
+    local: false,
+    creds: [
+      { key: 'AZURE_ANTHROPIC_ENDPOINT', label: 'Azure Anthropic endpoint URL' },
+      { key: 'AZURE_ANTHROPIC_API_KEY', label: 'Azure Anthropic API key', secret: true },
+    ],
+    extras: [],
+    defaultModel: 'claude-sonnet-4-6',
+  },
+  'azure-openai': {
+    label: 'Azure OpenAI (GPT family via Azure)',
+    local: false,
+    creds: [
+      { key: 'AZURE_OPENAI_ENDPOINT', label: 'Azure OpenAI endpoint URL' },
+      { key: 'AZURE_OPENAI_API_KEY', label: 'Azure OpenAI API key', secret: true },
+      { key: 'AZURE_OPENAI_DEPLOYMENT', label: 'Deployment name', default: 'gpt-5.2-chat' },
+    ],
+    extras: [],
+    defaultModel: 'gpt-5.2-chat',
+  },
+  openai: {
+    label: 'OpenAI (direct)',
+    local: false,
+    creds: [
+      { key: 'OPENAI_API_KEY', label: 'OpenAI API key', secret: true },
+    ],
+    extras: [],
+    defaultModel: 'gpt-4o',
+  },
+  openrouter: {
+    label: 'OpenRouter (100+ models, one key)',
+    local: false,
+    creds: [
+      { key: 'OPENROUTER_API_KEY', label: 'OpenRouter API key', secret: true },
+    ],
+    extras: [],
+    defaultModel: 'anthropic/claude-sonnet-4',
+  },
+  databricks: {
+    label: 'Databricks Foundation Models',
+    local: false,
+    creds: [
+      { key: 'DATABRICKS_API_BASE', label: 'Databricks workspace URL' },
+      { key: 'DATABRICKS_API_KEY', label: 'Databricks API token', secret: true },
+    ],
+    extras: [],
+    defaultModel: 'databricks-claude-sonnet-4',
+  },
+  bedrock: {
+    label: 'AWS Bedrock',
+    local: false,
+    creds: [
+      { key: 'BEDROCK_API_KEY', label: 'AWS Bedrock API key (or use IAM)', secret: true },
+    ],
+    extras: [],
+    defaultModel: 'anthropic.claude-sonnet-4-v1:0',
+  },
+  vertex: {
+    label: 'Google Vertex AI',
+    local: false,
+    creds: [
+      { key: 'VERTEX_API_KEY', label: 'Vertex API key (or use ADC)', secret: true },
+    ],
+    extras: [],
+    defaultModel: 'gemini-2.0-flash',
+  },
+  zai: {
+    label: 'Z.ai (GLM family)',
+    local: false,
+    creds: [
+      { key: 'ZAI_API_KEY', label: 'Z.ai API key', secret: true },
+    ],
+    extras: [],
+    defaultModel: 'GLM-4.7',
+  },
+  moonshot: {
+    label: 'Moonshot (Kimi family)',
+    local: false,
+    creds: [
+      { key: 'MOONSHOT_API_KEY', label: 'Moonshot API key', secret: true },
+    ],
+    extras: [],
+    defaultModel: 'kimi-k2-turbo-preview',
+  },
+};
+
+const PROVIDER_ORDER = [
+  'ollama', 'llamacpp', 'lmstudio',
+  'azure-anthropic', 'azure-openai', 'openai', 'openrouter',
+  'databricks', 'bedrock', 'vertex', 'zai', 'moonshot',
+];
+const TIERS = ['SIMPLE', 'MEDIUM', 'COMPLEX', 'REASONING'];
+
+// Always-emitted baseline. Mirrors the production-grade config the maintainer
+// runs locally: caching/compression on, generous policy budgets, sandboxed
+// agents, MCP/web defaults, etc. Users can edit any of these post-generation;
+// the wizard prompts only for tier picks, credentials, and a handful of
+// intelligence knobs. Everything else is opinionated default.
+//
+// Categorised inline so future contributors know which group a key lives in.
+const BASELINE_ENV = {
+  // ── Databricks placeholders (satisfy startup validator) ───────────────
+  DATABRICKS_API_BASE: 'http://localhost:8081',
+  DATABRICKS_API_KEY: 'tier-routing-active',
+  DATABRICKS_ENDPOINT_PATH: '/unused',
+
+  // ── Server ────────────────────────────────────────────────────────────
+  PORT: '8081',
+  NODE_ENV: 'production',
+  REQUEST_JSON_LIMIT: '1gb',
+  SESSION_DB_PATH: './data/sessions.db',
+  ENABLE_TOOL_SEARCH: 'true',
+  LOG_LEVEL: 'silent',
+
+  // ── Routing intelligence (tuned defaults) ─────────────────────────────
+  LYNKR_PREFLIGHT_ENABLED: 'false',
+  LYNKR_PREFLIGHT_TIMEOUT_MS: '120000',
+  LYNKR_CASCADE_ENABLED: 'true',
+  LYNKR_KNN_MIN_INDEX_SIZE: '200',
+  LYNKR_KNN_CONFIDENCE_HIGH: '0.55',
+  LYNKR_KNN_CONFIDENCE_LOW: '0.30',
+
+  // ── Tool execution ────────────────────────────────────────────────────
+  TOOL_EXECUTION_MODE: 'client',
+  TOOL_INJECTION_ENABLED: 'false',
+  SMART_TOOL_SELECTION_MODE: 'disabled',
+  SMART_TOOL_SELECTION_TOKEN_BUDGET: '2500',
+  CODE_MODE_ENABLED: 'true',
+
+  // ── Caching ───────────────────────────────────────────────────────────
+  PROMPT_CACHE_ENABLED: 'true',
+  PROMPT_CACHE_MAX_ENTRIES: '1000',
+  PROMPT_CACHE_TTL_MS: '300000',
+  SEMANTIC_CACHE_ENABLED: 'true',
+  SEMANTIC_CACHE_THRESHOLD: '0.85',
+  SEMANTIC_CACHE_MAX_ENTRIES: '50',
+  SEMANTIC_CACHE_TTL_MS: '300000',
+
+  // ── Compression: TOON + Headroom sidecar ──────────────────────────────
+  TOON_ENABLED: 'true',
+  TOON_MIN_BYTES: '4096',
+  TOON_FAIL_OPEN: 'true',
+  TOON_LOG_STATS: 'true',
+  HEADROOM_ENABLED: 'true',
+  HEADROOM_ENDPOINT: 'http://localhost:8787',
+  HEADROOM_TIMEOUT_MS: '5000',
+  HEADROOM_MIN_TOKENS: '100',
+  HEADROOM_MODE: 'optimize',
+  HEADROOM_PROVIDER: 'anthropic',
+  HEADROOM_DOCKER_ENABLED: 'true',
+  HEADROOM_DOCKER_IMAGE: 'lynkr/headroom-sidecar:latest',
+  HEADROOM_DOCKER_CONTAINER_NAME: 'lynkr-headroom',
+  HEADROOM_DOCKER_PORT: '8787',
+  HEADROOM_DOCKER_AUTO_BUILD: 'true',
+  HEADROOM_SMART_CRUSHER: 'true',
+  HEADROOM_SMART_CRUSHER_MIN_TOKENS: '200',
+  HEADROOM_SMART_CRUSHER_MAX_ITEMS: '15',
+  HEADROOM_TOOL_CRUSHER: 'true',
+  HEADROOM_CACHE_ALIGNER: 'true',
+  HEADROOM_ROLLING_WINDOW: 'true',
+  HEADROOM_KEEP_TURNS: '10',
+  HEADROOM_CCR: 'true',
+  HEADROOM_CCR_TTL: '300',
+
+  // ── Memory + token tracking ───────────────────────────────────────────
+  MEMORY_ENABLED: 'true',
+  MEMORY_RETRIEVAL_LIMIT: '5',
+  MEMORY_SURPRISE_THRESHOLD: '0.3',
+  MEMORY_MAX_AGE_DAYS: '90',
+  MEMORY_MAX_COUNT: '10000',
+  MEMORY_EXTRACTION_ENABLED: 'true',
+  MEMORY_DECAY_ENABLED: 'true',
+  MEMORY_DECAY_HALF_LIFE: '30',
+  MEMORY_FORMAT: 'compact',
+  MEMORY_DEDUP_ENABLED: 'true',
+  MEMORY_DEDUP_LOOKBACK: '5',
+  MEMORY_TTL: '3600000',
+  TOKEN_TRACKING_ENABLED: 'true',
+  TOOL_TRUNCATION_ENABLED: 'true',
+
+  // ── Prompt/output shaping ─────────────────────────────────────────────
+  SYSTEM_PROMPT_MODE: 'dynamic',
+  TOOL_DESCRIPTIONS: 'minimal',
+  HISTORY_COMPRESSION_ENABLED: 'true',
+  HISTORY_KEEP_RECENT_TURNS: '10',
+  HISTORY_SUMMARIZE_OLDER: 'true',
+  TOKEN_BUDGET_WARNING: '100000',
+  TOKEN_BUDGET_MAX: '180000',
+  TOKEN_BUDGET_ENFORCEMENT: 'true',
+  CAVEMAN_ENABLED: 'false',
+  CAVEMAN_LEVEL: 'full',
+  MARKDOWN_RENDER_ANSI: 'false',
+
+  // ── Policy & budgets ──────────────────────────────────────────────────
+  POLICY_MAX_STEPS: '2000',
+  POLICY_MAX_TOOL_CALLS: '2000',
+  POLICY_TOOL_LOOP_THRESHOLD: '100',
+  POLICY_GIT_ALLOW_PUSH: 'false',
+  POLICY_GIT_ALLOW_PULL: 'true',
+  POLICY_GIT_ALLOW_COMMIT: 'true',
+  POLICY_GIT_REQUIRE_TESTS: 'false',
+  POLICY_GIT_AUTOSTASH: 'false',
+  POLICY_FILE_BLOCKED_PATHS: '/.env,.env,/etc/passwd,/etc/shadow',
+  POLICY_SAFE_COMMANDS_ENABLED: 'true',
+
+  // ── Agents ────────────────────────────────────────────────────────────
+  AGENTS_ENABLED: 'true',
+  AGENTS_MAX_CONCURRENT: '10',
+  AGENTS_DEFAULT_MODEL: 'haiku',
+  AGENTS_MAX_STEPS: '15',
+  AGENTS_TIMEOUT: '300000',
+
+  // ── Rate limiting ─────────────────────────────────────────────────────
+  RATE_LIMIT_ENABLED: 'true',
+  RATE_LIMIT_WINDOW_MS: '60000',
+  RATE_LIMIT_MAX: '100',
+  RATE_LIMIT_KEY_BY: 'session',
+
+  // ── Hot reload + load shedding ────────────────────────────────────────
+  HOT_RELOAD_ENABLED: 'true',
+  HOT_RELOAD_DEBOUNCE_MS: '1000',
+  LOAD_SHEDDING_ENABLED: 'true',
+  LOAD_SHEDDING_HEAP_THRESHOLD: '0.85',
+  LOAD_SHEDDING_MEMORY_THRESHOLD: '0.95',
+
+  // ── Per-provider extras (secrets stay empty; wizard or user fills in) ─
+  AZURE_ANTHROPIC_ENDPOINT: 'https://api.anthropic.com/v1/messages',
+  AZURE_ANTHROPIC_VERSION: '2023-06-01',
+  AZURE_OPENAI_API_VERSION: '2024-08-01-preview',
+  OLLAMA_MODEL: 'minimax-m2.5:cloud',
+  OLLAMA_TIMEOUT_MS: '120000',
+  OLLAMA_EMBEDDINGS_MODEL: 'nomic-embed-text',
+  OLLAMA_EMBEDDINGS_ENDPOINT: 'http://localhost:11434/api/embeddings',
+  OPENROUTER_API_KEY: '',
+  OPENROUTER_MODEL: 'openai/gpt-4o-mini',
+  OPENROUTER_EMBEDDINGS_MODEL: 'openai/text-embedding-ada-002',
+  OPENROUTER_ENDPOINT: 'https://openrouter.ai/api/v1/chat/completions',
+  OPENROUTER_MAX_TOOLS_FOR_ROUTING: '15',
+  MOONSHOT_API_KEY: '',
+  MOONSHOT_ENDPOINT: 'https://api.moonshot.ai/v1/chat/completions',
+  MOONSHOT_MODEL: 'kimi-k2.6',
+  LLAMACPP_ENDPOINT: 'http://localhost:8080',
+  LLAMACPP_MODEL: 'default',
+  LLAMACPP_TIMEOUT_MS: '120000',
+  LLAMACPP_EMBEDDINGS_ENDPOINT: 'http://localhost:8080/embeddings',
+  LMSTUDIO_ENDPOINT: 'http://localhost:1234',
+  LMSTUDIO_MODEL: 'default',
+  LMSTUDIO_TIMEOUT_MS: '120000',
+
+  // ── MCP sandbox (Docker-isolated MCP tool execution) ──────────────────
+  MCP_SANDBOX_ENABLED: 'true',
+  MCP_SANDBOX_RUNTIME: 'docker',
+  MCP_SANDBOX_CONTAINER_WORKSPACE: '/workspace',
+  MCP_SANDBOX_MOUNT_WORKSPACE: 'true',
+  MCP_SANDBOX_ALLOW_NETWORKING: 'false',
+  MCP_SANDBOX_NETWORK_MODE: 'none',
+  MCP_SANDBOX_PASSTHROUGH_ENV: 'PATH,LANG,LC_ALL,TERM,HOME',
+  MCP_SANDBOX_TIMEOUT_MS: '20000',
+  MCP_SANDBOX_REUSE_SESSION: 'true',
+  MCP_SANDBOX_READ_ONLY_ROOT: 'false',
+  MCP_SANDBOX_NO_NEW_PRIVILEGES: 'true',
+  MCP_SANDBOX_DROP_CAPABILITIES: 'ALL',
+  MCP_SANDBOX_MEMORY_LIMIT: '512m',
+  MCP_SANDBOX_CPU_LIMIT: '1.0',
+  MCP_SANDBOX_PIDS_LIMIT: '100',
+  MCP_SANDBOX_PERMISSION_MODE: 'auto',
+  MCP_MANIFEST_DIRS: '~/.claude/mcp',
+
+  // ── Web tools (search + fetch) ────────────────────────────────────────
+  WEB_SEARCH_ENDPOINT: 'http://localhost:8888/search',
+  WEB_SEARCH_ALLOW_ALL: 'true',
+  WEB_SEARCH_TIMEOUT_MS: '10000',
+  WEB_FETCH_BODY_PREVIEW_MAX: '10000',
+  WEB_SEARCH_RETRY_ENABLED: 'true',
+  WEB_SEARCH_MAX_RETRIES: '2',
+
+  // ── TinyFish (web automation) ─────────────────────────────────────────
+  TINYFISH_API_KEY: '',
+  TINYFISH_ENDPOINT: 'https://agent.tinyfish.ai/v1/automation/run-sse',
+  TINYFISH_BROWSER_PROFILE: 'lite',
+  TINYFISH_TIMEOUT_MS: '120000',
+  TINYFISH_PROXY_ENABLED: 'false',
+  TINYFISH_PROXY_COUNTRY: 'US',
+
+  // ── Workspace test runner ─────────────────────────────────────────────
+  WORKSPACE_TEST_TIMEOUT_MS: '600000',
+  WORKSPACE_TEST_SANDBOX: 'auto',
+  WORKSPACE_TEST_COVERAGE_FILES: 'coverage/coverage-summary.json',
+};
+
+// ──────────────────────────────────────────────────────────────────────────────
+// Args
+// ──────────────────────────────────────────────────────────────────────────────
+
+function parseArgs(argv) {
+  const opts = { force: false, dryRun: false, output: null, help: false };
+  for (let i = 0; i < argv.length; i++) {
+    const a = argv[i];
+    if (a === '--help' || a === '-h') opts.help = true;
+    else if (a === '--force' || a === '-f') opts.force = true;
+    else if (a === '--dry-run') opts.dryRun = true;
+    else if (a.startsWith('--output=')) opts.output = a.slice('--output='.length);
+    else if (a === '--output' || a === '-o') opts.output = argv[++i];
+  }
+  return opts;
+}
+
+function showHelp() {
+  console.log(`lynkr init — interactive setup wizard
+
+Usage:
+  lynkr init                        Interactive wizard
+  lynkr init --force                Overwrite existing .env
+  lynkr init --output=<path>        Write to <path> instead of .env
+  lynkr init --dry-run              Print to stdout, don't write
+  lynkr init --help
+
+The wizard asks for:
+  1. Usage mode (Claude Pro/Max via wrap, or direct API keys)
+  2. Provider + model for each tier (SIMPLE / MEDIUM / COMPLEX / REASONING)
+  3. Credentials for each picked provider (re-used across tiers)
+  4. Routing intelligence (visible badge, intent window, decay)
+
+Providers covered: ${PROVIDER_ORDER.join(', ')}.
+`);
+}
+
+// ──────────────────────────────────────────────────────────────────────────────
+// Prompt helpers
+// ──────────────────────────────────────────────────────────────────────────────
+
+function makeAsker() {
+  const rl = readline.createInterface({ input: process.stdin, output: process.stdout });
+  const ask = (q) => new Promise((res) => rl.question(q, (a) => res(a.trim())));
+  const close = () => rl.close();
+  return { ask, close };
+}
+
+async function pickFromList(ask, label, choices, defaultIdx = 0) {
+  console.log(`\n${label}`);
+  choices.forEach((c, i) => {
+    const marker = i === defaultIdx ? '>' : ' ';
+    console.log(`  ${marker} ${i + 1}) ${c}`);
+  });
+  const raw = await ask(`Choice [1-${choices.length}] (default ${defaultIdx + 1}): `);
+  if (!raw) return defaultIdx;
+  const n = parseInt(raw, 10);
+  if (Number.isNaN(n) || n < 1 || n > choices.length) {
+    console.log(`  → invalid, using default (${choices[defaultIdx]})`);
+    return defaultIdx;
+  }
+  return n - 1;
+}
+
+async function askWithDefault(ask, label, defaultValue) {
+  const v = await ask(`${label}${defaultValue ? ` [${defaultValue}]` : ''}: `);
+  return v || defaultValue || '';
+}
+
+async function askYesNo(ask, label, defaultYes = true) {
+  const v = await ask(`${label} [${defaultYes ? 'Y/n' : 'y/N'}]: `);
+  if (!v) return defaultYes;
+  return /^y(es)?$/i.test(v);
+}
+
+// ──────────────────────────────────────────────────────────────────────────────
+// Wizard
+// ──────────────────────────────────────────────────────────────────────────────
+
+async function runInteractive(opts) {
+  console.log('lynkr init — interactive setup\n');
+  const { ask, close } = makeAsker();
+  const env = {};
+  const credsCollected = {}; // dedupe per env key
+
+  try {
+    // ── 1. Usage mode ──
+    const modeIdx = await pickFromList(ask,
+      'Usage mode:',
+      [
+        'Claude Pro/Max subscription (via `lynkr wrap claude`, OAuth passthrough)',
+        'Direct API usage (pay-as-you-go with API keys)',
+      ],
+      0,
+    );
+    const isWrap = modeIdx === 0;
+
+    if (isWrap) {
+      env.LYNKR_OAUTH_PASSTHROUGH = 'true';
+      console.log('\n  → OAuth passthrough enabled. COMPLEX/REASONING tiers will be sent');
+      console.log('    byte-for-byte to api.anthropic.com against your subscription.');
+      console.log('    You only need to configure a local model for SIMPLE/MEDIUM.\n');
+    }
+
+    // ── 2. Per-tier provider + model ──
+    const tierConfig = {};
+    const collectCreds = async (providerKey) => {
+      const p = PROVIDERS[providerKey];
+      for (const c of p.creds) {
+        if (credsCollected[c.key]) continue;
+        const existing = process.env[c.key];
+        const def = existing || c.default || '';
+        const prompt = `  ${c.label}${c.secret ? ' (hidden output not supported; paste anyway)' : ''}`;
+        const v = await askWithDefault(ask, prompt, def);
+        if (v) {
+          env[c.key] = v;
+          credsCollected[c.key] = true;
+        }
+      }
+      for (const ex of p.extras) {
+        if (env[ex.key]) continue;
+        const v = await askWithDefault(ask, `  ${ex.label}`, ex.default);
+        if (v) env[ex.key] = v;
+      }
+    };
+
+    const providerChoices = PROVIDER_ORDER.map((k) => PROVIDERS[k].label);
+
+    for (const tier of TIERS) {
+      const headline = isWrap && (tier === 'COMPLEX' || tier === 'REASONING')
+        ? `Tier ${tier} — covered by Pro/Max subscription, but you can override:`
+        : `Tier ${tier} — pick a provider:`;
+      const defaultIdx = isWrap && (tier === 'COMPLEX' || tier === 'REASONING')
+        ? PROVIDER_ORDER.indexOf('azure-anthropic')
+        : 0;
+
+      const skipOpt = isWrap && (tier === 'COMPLEX' || tier === 'REASONING')
+        ? [...providerChoices, 'Skip — let subscription passthrough handle it']
+        : providerChoices;
+
+      const idx = await pickFromList(ask, headline, skipOpt, defaultIdx);
+
+      if (idx === providerChoices.length) {
+        // Skip selected — leave TIER_<tier> unset
+        continue;
+      }
+
+      const providerKey = PROVIDER_ORDER[idx];
+      const p = PROVIDERS[providerKey];
+      const model = await askWithDefault(ask, `  Model for ${tier}`, p.defaultModel);
+      tierConfig[tier] = { provider: providerKey, model };
+      await collectCreds(providerKey);
+    }
+
+    for (const tier of TIERS) {
+      if (tierConfig[tier]) {
+        env[`TIER_${tier}`] = `${tierConfig[tier].provider}:${tierConfig[tier].model}`;
+      }
+    }
+
+    // Primary provider hint for legacy code paths
+    const firstTier = TIERS.map((t) => tierConfig[t]).find(Boolean);
+    if (firstTier) env.MODEL_PROVIDER = firstTier.provider;
+
+    // ── 3. Routing intelligence ──
+    console.log('\nRouting intelligence:');
+    if (await askYesNo(ask, 'Show the routing badge in your TUI (`*[Lynkr] …*`)?', isWrap)) {
+      env.LYNKR_VISIBLE_ROUTING = 'true';
+    }
+
+    const windowRaw = await askWithDefault(ask, 'Intent-scoring window size (1 = latest message only)', '5');
+    const windowN = parseInt(windowRaw, 10);
+    if (!Number.isNaN(windowN) && windowN >= 1) env.LYNKR_INTENT_WINDOW_N = String(windowN);
+
+    const decayRaw = await askWithDefault(ask, 'Intent-scoring per-turn decay (0.1-1.0)', '0.7');
+    const decay = parseFloat(decayRaw);
+    if (!Number.isNaN(decay) && decay > 0 && decay <= 1) env.LYNKR_INTENT_DECAY = String(decay);
+
+    close();
+    console.log('');
+    writeEnvFile(buildEnvContent(env, isWrap, tierConfig), opts);
+  } catch (err) {
+    close();
+    throw err;
+  }
+}
+
+// ──────────────────────────────────────────────────────────────────────────────
+// Output
+// ──────────────────────────────────────────────────────────────────────────────
+
+function buildEnvContent(env, isWrap, tierConfig) {
+  // Baseline first, user choices on top — so user input always wins for keys
+  // they explicitly answered (e.g. LOG_LEVEL if the wizard ever asks for it).
+  const merged = { ...BASELINE_ENV, ...env };
+
+  const lines = [
+    '# Lynkr configuration',
+    `# Generated by 'lynkr init' at ${new Date().toISOString()}`,
+    `# Mode: ${isWrap ? 'wrap (Claude Pro/Max subscription)' : 'direct API'}`,
+    '# Edit directly to tweak; full reference in .env.example',
+    '',
+  ];
+
+  // Group output by section in the order it appears in the generated file.
+  // Mirrors the layout of the .env.example reference doc.
+  const SERVER_KEYS = new Set(['PORT', 'NODE_ENV', 'REQUEST_JSON_LIMIT', 'SESSION_DB_PATH', 'WORKSPACE_ROOT', 'ENABLE_TOOL_SEARCH']);
+  const TOOL_EXEC_KEYS = new Set(['TOOL_EXECUTION_MODE', 'TOOL_INJECTION_ENABLED', 'SMART_TOOL_SELECTION_MODE', 'SMART_TOOL_SELECTION_TOKEN_BUDGET', 'CODE_MODE_ENABLED']);
+  const CACHE_KEYS = new Set([
+    'PROMPT_CACHE_ENABLED', 'PROMPT_CACHE_MAX_ENTRIES', 'PROMPT_CACHE_TTL_MS',
+    'SEMANTIC_CACHE_ENABLED', 'SEMANTIC_CACHE_THRESHOLD', 'SEMANTIC_CACHE_MAX_ENTRIES', 'SEMANTIC_CACHE_TTL_MS',
+  ]);
+  const MEMORY_KEYS = new Set(Object.keys(merged).filter((k) => k.startsWith('MEMORY_') || k === 'TOKEN_TRACKING_ENABLED' || k === 'TOOL_TRUNCATION_ENABLED'));
+  const SHAPING_KEYS = new Set([
+    'SYSTEM_PROMPT_MODE', 'TOOL_DESCRIPTIONS',
+    'HISTORY_COMPRESSION_ENABLED', 'HISTORY_KEEP_RECENT_TURNS', 'HISTORY_SUMMARIZE_OLDER',
+    'TOKEN_BUDGET_WARNING', 'TOKEN_BUDGET_MAX', 'TOKEN_BUDGET_ENFORCEMENT',
+    'CAVEMAN_ENABLED', 'CAVEMAN_LEVEL', 'MARKDOWN_RENDER_ANSI',
+  ]);
+  const POLICY_KEYS = new Set(Object.keys(merged).filter((k) => k.startsWith('POLICY_')));
+  const AGENT_KEYS = new Set(Object.keys(merged).filter((k) => k.startsWith('AGENTS_')));
+  const RATE_KEYS = new Set(Object.keys(merged).filter((k) => k.startsWith('RATE_LIMIT_')));
+  const OPS_KEYS = new Set(Object.keys(merged).filter((k) => k.startsWith('HOT_RELOAD_') || k.startsWith('LOAD_SHEDDING_')));
+  const COMPRESSION_KEYS = new Set(Object.keys(merged).filter((k) => k.startsWith('TOON_') || k.startsWith('HEADROOM_')));
+  const MCP_KEYS = new Set(Object.keys(merged).filter((k) => k.startsWith('MCP_')));
+  const WEB_KEYS = new Set(Object.keys(merged).filter((k) => k.startsWith('WEB_SEARCH_') || k.startsWith('WEB_FETCH_')));
+  const TINYFISH_KEYS = new Set(Object.keys(merged).filter((k) => k.startsWith('TINYFISH_')));
+  const WORKSPACE_TEST_KEYS = new Set(Object.keys(merged).filter((k) => k.startsWith('WORKSPACE_TEST_')));
+
+  const groups = [
+    { heading: '# Tier routing',           keys: Object.keys(merged).filter((k) => k.startsWith('TIER_') || k === 'MODEL_PROVIDER') },
+    { heading: '# Server',                 keys: Object.keys(merged).filter((k) => SERVER_KEYS.has(k)) },
+    { heading: '# Provider credentials',   keys: Object.keys(merged).filter((k) =>
+      /(_API_KEY|_ENDPOINT|_API_BASE|_DEPLOYMENT|_MODEL|_ENDPOINT_PATH|_API_VERSION|_VERSION|_TIMEOUT_MS|_EMBEDDINGS_MODEL|_EMBEDDINGS_ENDPOINT|_MAX_TOOLS_FOR_ROUTING)$/.test(k) &&
+      !k.startsWith('LYNKR_') && !k.startsWith('HEADROOM_') && !k.startsWith('RATE_LIMIT_') &&
+      !k.startsWith('HOT_RELOAD_') && !k.startsWith('LOAD_SHEDDING_') && !k.startsWith('AGENTS_') &&
+      !k.startsWith('MCP_') && !k.startsWith('WEB_') && !k.startsWith('TINYFISH_') && !k.startsWith('WORKSPACE_TEST_') &&
+      !k.startsWith('NODE_') && !k.startsWith('TOON_')
+    ) },
+    { heading: '# Routing intelligence',   keys: Object.keys(merged).filter((k) => k.startsWith('LYNKR_')) },
+    { heading: '# Tool execution',         keys: Object.keys(merged).filter((k) => TOOL_EXEC_KEYS.has(k)) },
+    { heading: '# Caching',                keys: Object.keys(merged).filter((k) => CACHE_KEYS.has(k)) },
+    { heading: '# Compression & context',  keys: Object.keys(merged).filter((k) => COMPRESSION_KEYS.has(k)) },
+    { heading: '# Memory & tracking',      keys: Object.keys(merged).filter((k) => MEMORY_KEYS.has(k)) },
+    { heading: '# Prompt & output shaping', keys: Object.keys(merged).filter((k) => SHAPING_KEYS.has(k)) },
+    { heading: '# Policy & budgets',       keys: Object.keys(merged).filter((k) => POLICY_KEYS.has(k)) },
+    { heading: '# Agents',                 keys: Object.keys(merged).filter((k) => AGENT_KEYS.has(k)) },
+    { heading: '# Rate limiting',          keys: Object.keys(merged).filter((k) => RATE_KEYS.has(k)) },
+    { heading: '# MCP sandbox',            keys: Object.keys(merged).filter((k) => MCP_KEYS.has(k)) },
+    { heading: '# Web tools',              keys: Object.keys(merged).filter((k) => WEB_KEYS.has(k)) },
+    { heading: '# TinyFish (web automation)', keys: Object.keys(merged).filter((k) => TINYFISH_KEYS.has(k)) },
+    { heading: '# Workspace test runner',  keys: Object.keys(merged).filter((k) => WORKSPACE_TEST_KEYS.has(k)) },
+    { heading: '# Ops (hot reload, load shedding)', keys: Object.keys(merged).filter((k) => OPS_KEYS.has(k)) },
+    { heading: '# Logging',                keys: ['LOG_LEVEL'].filter((k) => k in merged) },
+  ];
+
+  const seen = new Set();
+  for (const g of groups) {
+    if (!g.keys.length) continue;
+    lines.push(g.heading);
+    for (const k of g.keys) {
+      if (seen.has(k)) continue;
+      lines.push(`${k}=${merged[k]}`);
+      seen.add(k);
+    }
+    lines.push('');
+  }
+
+  // Catch-all for any other keys (e.g. _DEPLOYMENT defaults) we missed.
+  const remaining = Object.keys(merged).filter((k) => !seen.has(k));
+  if (remaining.length) {
+    lines.push('# Other');
+    for (const k of remaining) lines.push(`${k}=${merged[k]}`);
+    lines.push('');
+  }
+
+  return lines.join('\n');
+}
+
+function writeEnvFile(content, opts) {
+  if (opts.dryRun) {
+    process.stdout.write(content);
+    return;
+  }
+  const target = opts.output || path.join(process.cwd(), '.env');
+  if (fs.existsSync(target) && !opts.force) {
+    console.error(`✗ ${target} already exists. Use --force to overwrite, or --output=<path>.`);
+    process.exit(1);
+  }
+  fs.writeFileSync(target, content);
+  console.log(`✓ Wrote ${target}`);
+}
+
+// ──────────────────────────────────────────────────────────────────────────────
+// Entry
+// ──────────────────────────────────────────────────────────────────────────────
+
+async function main() {
+  const opts = parseArgs(process.argv.slice(2));
+  if (opts.help) return showHelp();
+
+  if (!process.stdin.isTTY) {
+    console.error('✗ lynkr init needs an interactive TTY.');
+    console.error('  If you need a non-interactive setup, copy .env.example to .env manually,');
+    console.error('  or run `lynkr init --dry-run` to preview the wizard prompts.');
+    process.exit(1);
+  }
+
+  return runInteractive(opts);
+}
+
+// Run when invoked directly (`node bin/lynkr-init.js`) or dispatched from
+// cli.js (which sets _LYNKR_SUBCMD). Stay quiet when require()'d by tests.
+if (require.main === module || process.env._LYNKR_SUBCMD === 'init') {
+  main().catch((err) => {
+    console.error(`✗ ${err.message}`);
+    process.exit(1);
+  });
+}
+
+module.exports = {
+  PROVIDERS,
+  PROVIDER_ORDER,
+  TIERS,
+  parseArgs,
+  buildEnvContent,
+};

package/install.sh

@@ -128,63 +128,17 @@ install_dependencies() {
     fi
 }
 
-# Create default .env file
+# Skip .env creation — the install script runs without a TTY when invoked via
+# `curl | bash`, so the interactive `lynkr init` wizard can't run here. We leave
+# .env unmade so the user is prompted to run `lynkr init` in their own shell
+# afterward, which produces a fully-populated config (~150 keys grouped by
+# section) instead of the old 892-line .env.example dump.
 create_env_file() {
-    if [ ! -f "$INSTALL_DIR/.env" ]; then
-        print_info "Creating .env configuration file..."
-
-        # Try to copy from .env.example (comprehensive configuration)
-        if [ -f "$INSTALL_DIR/.env.example" ]; then
-            cp "$INSTALL_DIR/.env.example" "$INSTALL_DIR/.env"
-            print_success "Created .env from .env.example (all features documented)"
-        else
-            # Fallback: create minimal .env if .env.example doesn't exist
-            cat > "$INSTALL_DIR/.env" << 'EOF'
-# Lynkr Configuration
-# For full options, see: https://github.com/Fast-Editor/Lynkr/blob/main/.env.example
-
-# Model Provider (databricks, openai, azure-openai, azure-anthropic, openrouter, ollama, llamacpp)
-MODEL_PROVIDER=ollama
-
-# Server Configuration
-PORT=8081
-
-# Ollama Configuration (default for local development)
-OLLAMA_MODEL=qwen2.5-coder:7b
-OLLAMA_ENDPOINT=http://localhost:11434
-
-# Tier-based routing (uncomment and configure to enable)
-# TIER_SIMPLE=ollama:qwen2.5-coder:7b
-# TIER_MEDIUM=ollama:qwen2.5-coder:7b
-# TIER_COMPLEX=ollama:qwen2.5-coder:7b
-# TIER_REASONING=ollama:qwen2.5-coder:7b
-
-# Long-Term Memory System (Titans-Inspired) - Enabled by default
-MEMORY_ENABLED=true
-MEMORY_RETRIEVAL_LIMIT=5
-MEMORY_SURPRISE_THRESHOLD=0.3
-
-# Uncomment and configure your preferred cloud provider:
-# OPENAI_API_KEY=sk-your-key
-# OPENROUTER_API_KEY=your-key
-# DATABRICKS_API_KEY=your-key
-# DATABRICKS_API_BASE=https://your-workspace.databricks.com
-EOF
-            print_success "Created basic .env file"
-        fi
-
-        echo ""
-        print_info "📝 Configuration ready! Key settings:"
-        echo "     • Default provider: Ollama (local, offline)"
-        echo "     • Memory system: Enabled (learns from conversations)"
-        echo "     • Port: 8081"
-        echo ""
-        print_warning "To use cloud providers (Databricks/OpenAI/Azure):"
-        echo "     Edit: ${BLUE}nano $INSTALL_DIR/.env${NC}"
-        echo "     Add your API keys and change MODEL_PROVIDER"
-    else
+    if [ -f "$INSTALL_DIR/.env" ]; then
         print_warning ".env file already exists, skipping"
+        return
     fi
+    print_info "Skipping .env creation — run ${BLUE}lynkr init${NC} after install for an interactive setup."
 }
 
 # Create symlink for global access
@@ -224,44 +178,35 @@ print_next_steps() {
     print_success "Lynkr installed successfully!"
     echo "=============================="
     echo ""
-    echo "🚀 Quick Start Guide:"
+    echo "🚀 Quick Start:"
     echo ""
-    echo "  ${GREEN}Option A: Use Ollama (Free, Local, Offline)${NC}"
-    echo "  ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━"
+    echo "  1. Run the setup wizard:"
+    echo "     ${BLUE}lynkr init${NC}  ${GREEN}← interactive config (4 prompts, ~30 sec)${NC}"
     echo ""
-    echo "  1. Install Ollama (if not already installed):"
-    echo "     ${BLUE}lynkr-setup${NC}  ${GREEN}← Automatic Ollama installer${NC}"
+    echo "     The wizard asks for your usage mode (Claude Pro/Max via wrap, or direct"
+    echo "     API), tier picks across 12 supported providers, credentials for what you"
+    echo "     chose, and a few routing knobs. It writes a fully-populated .env with"
+    echo "     production defaults for everything else (caching, compression, policy"
+    echo "     budgets, MCP sandbox, agents, rate limiting)."
     echo ""
     echo "  2. Start Lynkr:"
-    echo "     ${BLUE}lynkr${NC}"
+    echo "     ${BLUE}lynkr${NC}                ${GREEN}← run as a proxy server${NC}"
+    echo "     ${BLUE}lynkr wrap claude${NC}    ${GREEN}← OR launch a wrapped AI tool${NC}"
     echo ""
-    echo "  3. Configure Claude Code CLI:"
+    echo "  3. Point your tool at Lynkr:"
     echo "     ${BLUE}export ANTHROPIC_BASE_URL=http://localhost:8081${NC}"
+    echo "     ${BLUE}export ANTHROPIC_API_KEY=any-non-empty-value${NC}"
     echo "     ${BLUE}claude${NC}"
     echo ""
-    echo "  ${YELLOW}Option B: Use Cloud Providers (Databricks/OpenAI/Azure)${NC}"
-    echo "  ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━"
-    echo ""
-    echo "  1. Edit configuration file:"
-    echo "     ${BLUE}nano $INSTALL_DIR/.env${NC}"
-    echo ""
-    echo "     Update these lines:"
-    echo "     ${BLUE}MODEL_PROVIDER=databricks${NC}  ${GREEN}← Change from 'ollama'${NC}"
-    echo "     ${BLUE}DATABRICKS_API_KEY=dapi_xxxxx${NC}  ${GREEN}← Add your key${NC}"
-    echo "     ${BLUE}DATABRICKS_API_BASE=https://your-workspace.databricks.com${NC}"
-    echo ""
-    echo "  2. Start Lynkr:"
-    echo "     ${BLUE}lynkr${NC}"
-    echo ""
-    echo "  3. Configure Claude Code CLI:"
-    echo "     ${BLUE}export ANTHROPIC_BASE_URL=http://localhost:8081${NC}"
-    echo "     ${BLUE}export ANTHROPIC_API_KEY=any-non-empty-value${NC}  ${GREEN}← Placeholder${NC}"
-    echo "     ${BLUE}claude${NC}"
+    echo "  ${YELLOW}Manual configuration (alternative)${NC}"
+    echo "  ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━"
+    echo "     Copy ${BLUE}.env.example${NC} to ${BLUE}.env${NC} and edit by hand if you prefer."
+    echo "     The 892-line template documents every available knob."
     echo ""
     echo "━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━"
     echo ""
-    echo "💡 ${YELLOW}Tip:${NC} Memory system is enabled by default"
-    echo "   Lynkr remembers preferences and project context across sessions"
+    echo "💡 ${YELLOW}Tip:${NC} Memory system, prompt caching, and TOON compression are all on"
+    echo "   by default. The wizard's defaults match a production-grade Lynkr setup."
     echo ""
     echo "📚 Documentation: ${BLUE}https://github.com/Fast-Editor/Lynkr${NC}"
     echo "💬 Discord: ${BLUE}https://discord.gg/qF7DDxrX${NC}"

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "lynkr",
-  "version": "9.7.0",
+  "version": "9.7.1",
   "description": "Self-hosted LLM gateway and tier-routing proxy for Claude Code, Cursor, and Codex. Routes across Ollama, AWS Bedrock, OpenRouter, Databricks, Azure OpenAI, llama.cpp, and LM Studio with prompt caching, MCP tools, and 60-80% cost savings.",
   "main": "index.js",
   "bin": {
@@ -16,7 +16,7 @@
     "dev": "nodemon index.js",
     "lint": "eslint src index.js",
     "test": "npm run test:unit && npm run test:performance",
-    "test:unit": "DATABRICKS_API_KEY=test-key DATABRICKS_API_BASE=http://test.com node --test test/routing.test.js test/hybrid-routing-integration.test.js test/passthrough-mode.test.js test/openrouter-error-resilience.test.js test/format-conversion.test.js test/azure-openai-config.test.js test/azure-openai-format-conversion.test.js test/azure-openai-routing.test.js test/azure-openai-streaming.test.js test/azure-openai-error-resilience.test.js test/azure-openai-integration.test.js test/openai-integration.test.js test/toon-compression.test.js test/llamacpp-integration.test.js test/resilience.test.js test/telemetry-routing.test.js test/memory/store.test.js test/memory/surprise.test.js test/memory/extractor.test.js test/memory/search.test.js test/memory/retriever.test.js test/distill.test.js test/large-payload.test.js test/code-mode.test.js test/prompt-cache-injection.test.js test/risk-analyzer.test.js test/interaction-block.test.js test/preflight.test.js test/token-reduction.test.js test/session-affinity.test.js test/model-registry-cost.test.js test/task-decomposition.test.js test/output-format-guard.test.js test/tier-fallback.test.js test/wrap.test.js",
+    "test:unit": "DATABRICKS_API_KEY=test-key DATABRICKS_API_BASE=http://test.com node --test test/routing.test.js test/hybrid-routing-integration.test.js test/web-tools.test.js test/passthrough-mode.test.js test/openrouter-error-resilience.test.js test/format-conversion.test.js test/azure-openai-config.test.js test/azure-openai-format-conversion.test.js test/azure-openai-routing.test.js test/azure-openai-streaming.test.js test/azure-openai-error-resilience.test.js test/azure-openai-integration.test.js test/openai-integration.test.js test/toon-compression.test.js test/llamacpp-integration.test.js test/resilience.test.js test/telemetry-routing.test.js test/memory/store.test.js test/memory/surprise.test.js test/memory/extractor.test.js test/memory/search.test.js test/memory/retriever.test.js test/distill.test.js test/large-payload.test.js test/code-mode.test.js test/prompt-cache-injection.test.js test/risk-analyzer.test.js test/interaction-block.test.js test/preflight.test.js test/token-reduction.test.js test/session-affinity.test.js test/model-registry-cost.test.js test/task-decomposition.test.js test/output-format-guard.test.js test/tier-fallback.test.js test/wrap.test.js test/init.test.js",
     "test:memory": "DATABRICKS_API_KEY=test-key DATABRICKS_API_BASE=http://test.com node --test test/memory/store.test.js test/memory/surprise.test.js test/memory/extractor.test.js test/memory/search.test.js test/memory/retriever.test.js",
     "test:new-features": "DATABRICKS_API_KEY=test-key DATABRICKS_API_BASE=http://test.com node --test test/passthrough-mode.test.js test/openrouter-error-resilience.test.js test/format-conversion.test.js",
     "test:performance": "DATABRICKS_API_KEY=test-key DATABRICKS_API_BASE=http://test.com node test/hybrid-routing-performance.test.js && DATABRICKS_API_KEY=test-key DATABRICKS_API_BASE=http://test.com node test/performance-tests.js",