memara 0.2.0

Official Python SDK for the Memara API - Give your AI a perfect memory

Memara Python SDK

The official Python SDK for the Memara API - Give your AI a perfect memory.

🚀 Quick Start

Installation

pip install memara

✅ Now Available on PyPI! Install with confidence - the package is live and ready to use.

Basic Usage

from memara import Memara

# Initialize the client
client = Memara(api_key="your_api_key_here")

# Create a memory
memory = client.create_memory(
    content="Important meeting notes from today",
    tags=["work", "meeting", "important"],
    importance=8
)

# Search memories
results = client.search_memories(
    query="meeting notes",
    limit=10
)

# List all memories
memories = client.list_memories(page=1, size=20)

# Close the client when done
client.close()

# Or use as a context manager (recommended)
with Memara(api_key="your_api_key") as client:
    memory = client.create_memory("Hello from Python SDK!")

📖 Documentation

Authentication

Get your API key from Memara Dashboard and set it either:

1. As environment variable (recommended):

export MEMARA_API_KEY="your_api_key_here"

2. Or pass directly:

client = Memara(api_key="your_api_key_here")

Configuration

# Full configuration options
client = Memara(
    api_key="your_api_key",           # Your Memara API key
    base_url="https://api.memara.io", # API base URL (optional)
    timeout=30.0                      # Request timeout in seconds
)

Memory Operations

Create Memory

memory = client.create_memory(
    content="The content of your memory",
    tags=["tag1", "tag2"],           # Optional: list of tags
    source="my_app",                 # Optional: source identifier  
    importance=7,                    # Optional: 1-10 importance level
    space_id="space_uuid"            # Optional: specific space ID
)

Search Memories

# Basic search
results = client.search_memories("your search query")

# Advanced search
results = client.search_memories(
    query="meeting notes",
    limit=20,                        # Max results to return
    space_id="specific_space_id",    # Search within specific space
    cross_space=False               # Search across all spaces
)

Get Memory by ID

memory = client.get_memory("memory_uuid")

# With space context
memory = client.get_memory("memory_uuid", space_id="space_uuid")

Delete Memory

result = client.delete_memory("memory_uuid")

Create Audio Memory (NEW in v0.2.0)

# Upload audio file with automatic transcription
memory = client.create_audio_memory(
    audio_file="meeting.mp3",           # Path to audio file
    content="Team standup meeting",     # Memory title/description
    tags=["meeting", "team"],           # Optional: tags
    importance=7,                       # Optional: importance (1-10)
    category="audio",                   # Optional: category
    space_id="space_uuid"               # Optional: specific space
)

# Access transcription and audio data
print(f"Transcription: {memory.metadata['audio_transcription']}")
print(f"Audio URL: {memory.metadata['audio_url']}")
print(f"Duration: {memory.metadata['audio_metadata']['duration_seconds']}s")

# Supported formats: MP3, M4A, WAV, FLAC, OGG, AAC
# Three input types supported:
#   1. File path (str): "path/to/audio.mp3"
#   2. Path object: Path("audio.mp3")
#   3. Raw bytes: audio_bytes

Audio Memory Features:

• 🎙️ Automatic Transcription: OpenAI Whisper transcription
• 🔍 Searchable: Audio transcriptions are fully searchable
• 📁 Secure Storage: Audio stored in S3 with CDN delivery
• 📊 Metadata: Duration, format, language, confidence scores
• 🎚️ Tier Limits: Respects tier-based file size limits

Space Operations

List Spaces

spaces = client.list_spaces()
for space in spaces:
    print(f"Space: {space.name} ({space.memory_count} memories)")

Create Space

space = client.create_space(
    name="My Project Space",
    icon="🚀",                      # Optional: emoji icon
    color="#6366F1",               # Optional: hex color
    template_type="work"           # Optional: template type
)

🔧 Advanced Usage

Error Handling

from memara import Memara, MemaraAPIError, MemaraAuthError

try:
    with Memara() as client:
        memory = client.create_memory("Test memory")
except MemaraAuthError:
    print("Authentication failed - check your API key")
except MemaraAPIError as e:
    print(f"API error: {e}")
except Exception as e:
    print(f"Unexpected error: {e}")

Environment Variables

Set these environment variables for easier configuration:

export MEMARA_API_KEY="your_api_key_here"
export MEMARA_API_URL="https://api.memara.io"  # Optional: custom API URL

Async Usage (Coming Soon)

Future versions will include async support:

# Coming in v0.2.0
from memara import AsyncMemara

async with AsyncMemara(api_key="your_key") as client:
    memory = await client.create_memory("Async memory!")

🛠️ Development

Contributing

1. Clone the repository
2. Install development dependencies: pip install -e .[dev]
3. Run tests: pytest
4. Format code: black memara/
5. Type check: mypy memara/

Running Tests

# Install with dev dependencies
pip install -e .[dev]

# Run tests
pytest

# Run tests with coverage
pytest --cov=memara

🔗 Links

• PyPI Package: pypi.org/project/memara 📦
• Homepage: memara.io
• Documentation: memara.io/docs
• API Reference: memara.io/docs/api
• GitHub: github.com/memara-ai/memara-python-sdk

📝 License

This project is licensed under the MIT License - see the LICENSE file for details.

🆘 Support

• Documentation: memara.io/docs
• Discord: discord.memara.io
• Email: [e-mail]
• Issues: GitHub Issues

---

Give your AI a perfect memory with Memara 🧠✨