tuningengines-cli 0.2.0 → 0.2.1

package/README.md

@@ -0,0 +1,206 @@
+# Tuning Engines CLI & MCP Server
+
+Fine-tune large language models from your terminal or AI assistant. Train custom models on your code repositories with a single command.
+
+**[Tuning Engines](https://tuningengines.com)** is a managed fine-tuning platform that handles GPU infrastructure, training orchestration, billing, and model storage — so you can focus on building better models.
+
+## Features
+
+- **Browser-based auth** — `te auth login` opens your browser to sign up or log in (like `gh auth login`)
+- **Full training lifecycle** — create, monitor, cancel, and retry fine-tuning jobs
+- **Model management** — list, import, export, and delete your trained models
+- **Cost estimation** — get estimates before committing to a training run
+- **S3 integration** — validate credentials and export models to your own S3 buckets
+- **MCP server** — 18 tools for AI assistants (Claude, ChatGPT, Cursor, VS Code Copilot, etc.)
+- **Billing** — check balance, open browser to add credits
+
+## Quick Start
+
+```bash
+npm install -g tuningengines-cli
+
+# Sign up or log in (opens browser)
+te auth login
+
+# Check your balance
+te billing show
+
+# Estimate cost for a training job
+te jobs estimate --base-model meta-llama/Llama-3.1-8B --repo-url https://github.com/you/repo
+
+# Create a training job
+te jobs create --base-model meta-llama/Llama-3.1-8B \
+  --repo-url https://github.com/you/repo \
+  --output-name my-model
+
+# Watch training progress
+te jobs status <job-id> --watch
+
+# List your trained models
+te models list
+```
+
+## MCP Server Setup
+
+The CLI includes a built-in MCP server with 18 tools for AI-powered fine-tuning workflows.
+
+### Claude Desktop
+
+Add to `~/Library/Application Support/Claude/claude_desktop_config.json`:
+
+```json
+{
+  "mcpServers": {
+    "tuning-engines": {
+      "command": "npx",
+      "args": ["-y", "tuningengines-cli", "mcp", "serve"],
+      "env": {
+        "TE_API_KEY": "te_your_key_here"
+      }
+    }
+  }
+}
+```
+
+### Claude Code
+
+```bash
+claude mcp add tuning-engines -- npx -y tuningengines-cli mcp serve
+```
+
+### VS Code / Cursor / Windsurf
+
+Add to your MCP settings (`.vscode/mcp.json` or equivalent):
+
+```json
+{
+  "servers": {
+    "tuning-engines": {
+      "command": "npx",
+      "args": ["-y", "tuningengines-cli", "mcp", "serve"],
+      "env": {
+        "TE_API_KEY": "te_your_key_here"
+      }
+    }
+  }
+}
+```
+
+### ChatGPT
+
+ChatGPT supports remote MCP servers. Add via Settings > Connectors with the Tuning Engines MCP URL.
+
+## CLI Commands
+
+### Authentication
+
+| Command | Description |
+|---------|-------------|
+| `te auth login` | Sign up or log in via browser |
+| `te auth logout` | Clear saved credentials |
+| `te auth status` | Show current auth status |
+
+### Training Jobs
+
+| Command | Description |
+|---------|-------------|
+| `te jobs list` | List all training jobs |
+| `te jobs show <id>` | Show job details |
+| `te jobs create` | Submit a new training job |
+| `te jobs status <id>` | Live job status (supports `--watch`) |
+| `te jobs cancel <id>` | Cancel a running job |
+| `te jobs retry <id>` | Retry from last checkpoint |
+| `te jobs estimate` | Get cost estimate |
+| `te jobs validate-s3` | Pre-validate S3 credentials |
+
+### Models
+
+| Command | Description |
+|---------|-------------|
+| `te models list` | List your trained models |
+| `te models show <id>` | Show model details |
+| `te models base` | List supported base models |
+| `te models import` | Import a model from S3 |
+| `te models export <id>` | Export a model to S3 |
+| `te models delete <id>` | Delete a model |
+| `te models status <id>` | Check import/export status |
+
+### Billing & Account
+
+| Command | Description |
+|---------|-------------|
+| `te billing show` | Show balance and transactions |
+| `te billing add-credits` | Open browser to add credits |
+| `te account` | Show account info |
+
+### Configuration
+
+| Command | Description |
+|---------|-------------|
+| `te config set-token <key>` | Set API key manually |
+| `te config set-url <url>` | Override API URL |
+| `te config show` | Show current config |
+
+All commands support `--json` for machine-readable output.
+
+## MCP Tools
+
+The MCP server exposes these tools for AI assistants:
+
+| Tool | Description |
+|------|-------------|
+| `list_jobs` | List training jobs with optional status filter |
+| `show_job` | Get details of a specific training job |
+| `create_job` | Submit a new fine-tuning job (supports `agent`, `quality_tier`) |
+| `cancel_job` | Cancel a running or queued job |
+| `job_status` | Get live status with GPU usage and charges |
+| `retry_job` | Retry a failed job from its last checkpoint |
+| `estimate_job` | Get cost estimate before submitting |
+| `validate_s3` | Validate S3 credentials before job submission |
+| `list_models` | List your trained and imported models |
+| `show_model` | Get details of a specific model |
+| `delete_model` | Delete a model from cloud storage |
+| `import_model` | Import a model from S3 |
+| `export_model` | Export a model to S3 |
+| `model_status` | Check model import/export status |
+| `list_supported_models` | List available base models for fine-tuning |
+| `get_balance` | Check account balance and transactions |
+| `get_account` | Get account information |
+
+### Example Prompts
+
+- "Fine-tune Llama 3.1 8B on my-org/my-repo and name it custom-model"
+- "How much would 3 epochs of training cost for an 8B model?"
+- "What's the status of my latest training job?"
+- "Show my account balance"
+- "List all my trained models"
+- "Export my model to s3://my-bucket/models/"
+
+## Environment Variables
+
+| Variable | Description |
+|----------|-------------|
+| `TE_API_KEY` | API key (overrides config file) |
+| `TE_API_URL` | API URL (default: `https://app.tuningengines.com`) |
+
+## Authentication
+
+`te auth login` uses a secure device authorization flow:
+
+1. CLI generates a device code and opens your browser
+2. You sign up or log in (email, Google, or GitHub)
+3. Click "Authorize" to grant CLI access
+4. Token flows back to CLI automatically — no copy-paste needed
+
+Token is saved to `~/.tuningengines/config.json` with `0600` permissions.
+
+## Links
+
+- [Website](https://tuningengines.com)
+- [Documentation](https://tuningengines.com)
+- [GitHub](https://github.com/ockhamlabs/tuning-engines-cli)
+- [npm](https://www.npmjs.com/package/tuningengines-cli)
+
+## License
+
+MIT

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "tuningengines-cli",
-  "version": "0.2.0",
-  "description": "CLI for Tuning Engines — fine-tune LLMs from your terminal",
+  "version": "0.2.1",
+  "description": "CLI and MCP server for Tuning Engines — fine-tune LLMs from your terminal or AI assistant. Train custom models on code repos with browser-based auth, cost estimation, model management, and 18 MCP tools for Claude, ChatGPT, Cursor, and more.",
   "main": "dist/cli.js",
   "bin": {
     "te": "dist/cli.js"
@@ -23,7 +23,8 @@
     "node": ">=18.0.0"
   },
   "files": [
-    "dist"
+    "dist",
+    "README.md"
   ],
   "license": "MIT",
   "mcpName": "io.github.cerebrixos/tuning-engines",
@@ -33,12 +34,20 @@
   },
   "keywords": [
     "mcp",
+    "mcp-server",
+    "model-context-protocol",
     "llm",
     "fine-tuning",
+    "fine-tune",
     "training",
     "ai",
+    "claude",
+    "chatgpt",
+    "cursor",
     "tuning-engines",
-    "model-context-protocol"
+    "machine-learning",
+    "lora",
+    "llama"
   ],
   "homepage": "https://tuningengines.com"
 }