python-max-bot 0.2.0

Async Python client for the Max messenger Bot API

python-max-bot

Async Python client for the Max messenger Bot API.

A thin, stable, Pydantic-modeled wrapper. No bot framework, no FSM, no middleware — just the API surface you need to send messages, upload attachments, and poll for updates.

Install

pip install python-max-bot

Quickstart

import asyncio
from max_bot_api import MaxClient

async def main():
    async with MaxClient(token="YOUR_BOT_TOKEN") as client:
        # Send a plain message
        await client.send_message(chat_id=42, text="Hello, Max!")

        # Send a message with an image
        with open("photo.jpg", "rb") as f:
            img = await client.upload_image(f.read(), filename="photo.jpg")
        await client.send_message(
            chat_id=42,
            text="<b>Look at this</b>",
            format="html",
            attachments=[img],
        )

asyncio.run(main())

What's in v0.1

Feature	Status
Send / edit / delete messages	✅
Long-poll updates (get_updates)	✅
Image / video / audio / file uploads	✅
Inline keyboards	✅
Chat metadata (get_chat)	✅
HTML and Markdown formatting	✅
Typed exceptions per HTTP status	✅
Auto-retry / backoff	✅
Webhook subscription endpoints	⏳ v0.3+
Bot framework (handlers, FSM)	❌ out of scope

Errors

from max_bot_api import (
    MaxClient,
    MaxAuthError,
    MaxRateLimitError,
    MaxTransportError,
)

async with MaxClient(token=...) as client:
    try:
        await client.send_message(chat_id=42, text="hi")
    except MaxAuthError:
        # Bad token — surface to user
        ...
    except MaxRateLimitError as e:
        await asyncio.sleep(e.retry_after or 1)
    except MaxTransportError:
        # Network problem; retry your way
        ...

Retries

Retries are opt-in. Pass a RetryPolicy to the client constructor:

from max_bot_api import MaxClient, RetryPolicy

async with MaxClient(token, retry=RetryPolicy()) as client:
    await client.get_chat(42)               # retries on 5xx and transport errors
    await client.send_message(chat_id=42, text="hi")  # only retries on transport errors

The default policy: 3 attempts, exponential backoff (1s, 2s, 4s with ±25% jitter), capped at 30s per wait. Tune via RetryPolicy(max_attempts=..., backoff_initial=..., backoff_multiplier=..., backoff_max=..., jitter=...).

Read methods (get_messages, get_updates, get_chat, request_upload_url) retry on both transport errors and 5xx responses. Write methods (send_message, edit_message, delete_message, the upload POST) only retry on transport errors — a 5xx during a write could mean the server processed it, so blind retry could double-apply.

429 responses always retry. The library waits at least as long as the server's Retry-After header asks; backoff_max does not clamp the server's instruction.

Without retry=, behavior is identical to v0.1 — one attempt per call.

Built with Claude

This project is developed in collaboration with Claude, Anthropic's AI assistant — design, code, and tests. Commits are co-authored. The repo ships with a CLAUDE.md and .claude/settings.json so any contributor running Claude on the codebase picks up the same conventions automatically. See the v0.1 design for the full collaboration workflow.

Links

• Max Bot API docs — upstream reference
• Design doc — what's in scope and what isn't
• CHANGELOG
• Contributing

License

MIT — see LICENSE.