mantr 1.0.0

Official Python SDK for Mantr - Deterministic Semantic Memory

Mantr Python SDK

Official Python client for Mantr - Deterministic Semantic Memory for AI.

Installation

pip install mantr

Quick Start

from mantr import MantrClient

# Initialize with your API key
client = MantrClient(api_key='vak_live_...')

# Walk the semantic graph
result = client.walk(phonemes=['dharma', 'karma', 'yoga'])

# Access results
for path in result.paths:
    print(f"{' → '.join(path.nodes)} (score: {path.score})")

Features

• ✅ Type-safe - Full Pydantic model validation
• ✅ Async support - Built-in retry logic with exponential backoff
• ✅ Error handling - Detailed exception types
• ✅ Context manager - Automatic cleanup
• ✅ No dependencies - Only requests and pydantic

API Reference

MantrClient

client = MantrClient(
    api_key: str,              # Your Mantr API key (required)
    base_url: str = "https://api.mantr.net",  # API endpoint
    timeout: int = 30          # Request timeout in seconds
)

walk()

Traverse the semantic graph to discover connections.

result = client.walk(
    phonemes: List[str],       # Sanskrit concepts to explore
    pod: Optional[str] = None, # Context pod (coming soon)
    depth: int = 3,            # Traversal depth (1-10)
    limit: int = 100           # Max results (1-1000)
)

Returns: WalkResponse with:

• paths - List of discovered paths
• latency_us - Execution time in microseconds
• credits_used - Credits consumed

Examples

Chatbot with Memory

from mantr import MantrClient

client = MantrClient(api_key='vak_live_...')

# Store conversation context
user_message = "How do I implement OAuth?"
context = client.walk(phonemes=['authentication', 'security'])

# Pass to your LLM
response = openai.chat.completions.create(
    model="gpt-4",
    messages=[
        {"role": "system", "content": f"Context: {context}"},
        {"role": "user", "content": user_message}
    ]
)

Document Q&A

# Query your knowledge base
answer = client.walk(
    phonemes=['pricing', 'policy'],
    depth=5,
    limit=10
)

for path in answer.paths:
    print(f"Relevance: {path.score:.2%} - {path.nodes}")

Error Handling

from mantr import (
    MantrClient,
    AuthenticationError,
    InsufficientCreditsError,
    RateLimitError
)

try:
    result = client.walk(['test'])
except AuthenticationError:
    print("Invalid API key")
except InsufficientCreditsError:
    print("Out of credits - upgrade your plan")
except RateLimitError as e:
    print(f"Rate limit exceeded - retry after {e.retry_after}s")

Context Manager

with MantrClient(api_key='vak_live_...') as client:
    result = client.walk(['dharma'])
# Session automatically closed

Get API Key

1. Sign up at mantr.net/signup
2. Get 5,000 free walks/month
3. No credit card required

Documentation

• API Docs: https://docs.mantr.net
• Examples: /examples
• Changelog: CHANGELOG.md

Support

• Issues: GitHub Issues
• Email: [e-mail]
• Discord: https://discord.gg/mantr

License

MIT License - see LICENSE for details.