# STIT — AI Garment Engineering OS
> STIT is the world's first agentic techpack platform for fashion brands, factories, and AI agents.
> Agents can autonomously create, manage, and export complete technical fashion packages (techpacks)
> via REST API, Python SDK, or MCP server integration.

## What is a Techpack?
A techpack is a complete technical specification document for manufacturing a garment. It includes:
- Bill of Materials (BOM): fabrics, trims, accessories with quantities and costs
- Spec Measurements: Points of Measure (POM) across all sizes with tolerance ranges
- Tech Notes: construction specifications, seam types, closures, finishing details
- Fitting Records: session notes, issues, audio transcriptions, quality ratings
- Design Annotations: canvas pins, comments, revision history
- Export: multilingual PDF/Excel for factory production

## API Access

### Base URL
Production: https://api.stit.ai/api/v1
Sandbox (free, no credits): https://sandbox.api.stit.ai/api/v1

### Authentication
```
POST /api-keys
Authorization: Bearer <jwt_token>
→ Returns: {key: "tp_sk_<64-hex>"}  (shown only once — store securely)

All subsequent requests:
Authorization: Bearer tp_sk_<your_key>
```

### OpenAPI Specification
https://api.stit.ai/openapi.json

### Interactive Docs
https://api.stit.ai/docs

## Integration Options

### MCP Server (Claude / Cursor / Cline)
```
npx @stit/mcp-server
```
Configure with your API key. Provides 15 tools:
stit_create_techpack, stit_generate_bom, stit_create_size_spec,
stit_add_pom_point, stit_add_measurement, stit_generate_grading,
stit_generate_tech_notes, stit_export_techpack, stit_poll_export,
stit_get_account_info, stit_list_techpacks, stit_get_techpack,
stit_advance_stage, stit_add_bom_item, stit_list_bom

### Python SDK
```
pip install stit-sdk
```
```python
from stit_sdk import STITSandboxClient
async with STITSandboxClient() as client:
    techpack = await client.techpacks.create(
        style_number="SS25-001", style_name="CELESTIA GOWN",
        category="dress", season="SS25"
    )
    bom = await client.bom.generate(techpack_id=techpack.id,
        description="Evening gown, polyester blend, geometric print")
    url = await client.exports.create_and_wait(techpack_id=techpack.id,
        format="PDF", language="en")
```

### TypeScript SDK
```
npm install @stit/sdk
```

## Core Capabilities

### Techpack Lifecycle
- Stages: Design → Proto → SMS → PPS → PP → Bulk
- Autonomy Levels: L1 (read-only suggestions) to L5 (full batch automation)
- Brand-factory pairing and collaboration
- Version control with full audit trail

### AI-Powered Generation (13 internal agents)
- BOM generation from garment description (confidence-scored)
- Tech notes generation from design context
- POM template inference by garment category
- Size grading table calculation (US/EU/CN standards)
- Fitting issue classification from audio transcripts
- PDF import extraction (spec sheets, existing techpacks)

### Export
- Formats: PDF, Excel (XLSX), CSV
- Languages: English, Chinese (Simplified/Traditional), Japanese, Korean,
  French, German, Spanish, Italian, Portuguese, Cantonese (11 total)
- Sections: BOM, Spec Measurements, Tech Notes, Fitting Records, Design
- Async generation with polling or Server-Sent Events

## Quick Start for Agents

```python
# 1. Get an API key (or use sandbox for free testing)
# POST https://api.stit.ai/api/v1/api-keys (requires user account)
# Sandbox key: pre-configured in STITSandboxClient

# 2. Create a techpack
POST /api/v1/techpacks
{"style_number": "SS25-001", "style_name": "My Garment", "category": "dress", "season": "SS25"}
→ {"id": "tp_...", "status": "design"}

# 3. Generate BOM (AI)
POST /api/v1/techpacks/{id}/bom/generate
{"description": "Evening gown, polyester blend, geometric print"}
→ [{item_code, item_name, category, quantity, unit_price, confidence}]

# 4. Add spec measurements
POST /api/v1/techpacks/{id}/spec/sizes
{"size_category": "US", "available_sizes": ["XS","S","M","L","XL"], "sample_size": "M"}

# 5. Export
POST /api/v1/export/techpack/{id}
{"format": "PDF", "language": "en", "quality": "print"}
→ {"export_id": "exp_...", "status": "pending"}

GET /api/v1/export/{export_id}/status
→ {"status": "completed", "download_url": "https://..."}
```

## Credit System
- Free tier: 200 credits on signup (no card required)
- Agent plan: 10,000 credits/month
- AI operations: 10–100 credits each (BOM generation ~50, export ~20)
- Insufficient credits: HTTP 402 with details
- Check balance: GET /api/v1/credits/balance

## Rate Limits
All responses include: RateLimit-Limit, RateLimit-Remaining, RateLimit-Reset headers
On 429: Retry-After header with seconds to wait

## Webhooks (Agent Callbacks)
Subscribe to events for async notification:
POST /api/v1/webhooks
{"url": "https://your-agent/callback", "events": ["export.completed", "techpack.status_changed"]}
Delivery: HMAC-SHA256 signed POST to your URL

## Pricing
https://stit.ai/pricing

## Developer Resources
- Agent Manifest: https://stit.ai/agent-manifest.json
- GitHub: https://github.com/stit-ai
- Docs: https://docs.stit.ai
- Support: api@stit.ai