Metadata-Version: 2.4
Name: agentraas
Version: 0.3.0
Summary: Exactly-once execution for AI agents: a local @exactly_once decorator (no server) and the AgentRaaS client
Author: Sumedh Chatse
License: MIT License
        
        Copyright (c) 2026 Sumedh Chatse
        
        Permission is hereby granted, free of charge, to any person obtaining a copy
        of this software and associated documentation files (the "Software"), to deal
        in the Software without restriction, including without limitation the rights
        to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
        copies of the Software, and to permit persons to whom the Software is
        furnished to do so, subject to the following conditions:
        
        The above copyright notice and this permission notice shall be included in all
        copies or substantial portions of the Software.
        
        THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
        IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
        FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
        AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
        LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
        OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
        SOFTWARE.
        
Project-URL: Homepage, https://github.com/sumedhchatse/agentraas
Project-URL: Repository, https://github.com/sumedhchatse/agentraas
Project-URL: Documentation, https://github.com/sumedhchatse/agentraas#readme
Project-URL: Issues, https://github.com/sumedhchatse/agentraas/issues
Keywords: exactly-once,deduplication,ai-agents,idempotency,mcp,webhooks,n8n,reliability
Classifier: Development Status :: 4 - Beta
Classifier: Intended Audience :: Developers
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.8
Classifier: Programming Language :: Python :: 3.9
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: Topic :: Software Development :: Libraries :: Python Modules
Requires-Python: >=3.8
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: requests>=2.28.0
Provides-Extra: redis
Requires-Dist: redis>=4.0; extra == "redis"
Dynamic: license-file

# agentraas (Python SDK)

Exactly-once execution for AI agents — a thin, dependency-light wrapper
around [AgentRaaS](https://github.com/sumedhchatse/agentraas)'s SDK-style
REST gateway. If your agent code calls Stripe, Twilio, HubSpot, or any
other API directly, wrap it with this client so a retry — yours, your
framework's, or a flaky network — never becomes a duplicate charge, a
duplicate contact, or a duplicate message.

Works against any AgentRaaS deployment: self-hosted (`./install.sh`,
free and unlimited on every tier) or AgentRaaS Cloud.

## Install

```bash
pip install agentraas
```

## Library mode: no server needed

Wrap any function that has a side effect (charging a card, sending an
email, creating a CRM record) so retries and concurrent duplicates run it
**exactly once**. Nothing to deploy: state lives in a local SQLite file by
default, or in your own Redis.

```python
from agentraas.local import exactly_once, RedisStore

@exactly_once()                                   # SQLite file, one machine
def charge(customer, amount):
    return stripe.Charge.create(customer=customer, amount=amount).id

charge("cus_1", 4200)   # runs for real
charge("cus_1", 4200)   # retry: returns the cached id, no second charge

# Across machines, with your own idempotency key and window:
@exactly_once(store=RedisStore("redis://localhost:6379/0"),
              key=lambda order: order["id"], ttl=3600)
async def ship(order):
    ...
```

- If the function raises, nothing is cached and a retry runs it again.
- A duplicate that arrives while the first call is still running waits for
  its result (up to `wait=30` seconds, then `InFlightError`).
- Results must be JSON-serializable (return ids or dicts, not SDK objects).
- `RedisStore` needs `pip install redis`.

When you want a dashboard, audit trail, human approval or circuit
breaking on top, point the same calls at an AgentRaaS server with the
`Client` below.

## Quickstart

1. Connect an agent from your AgentRaaS dashboard (**+ Connect Agent**) —
   this gives you an `agentraas_key`, plus your `org_id` and `agent_id`.
2. Add credentials for the service you're calling (**Credentials** panel)
   — AgentRaaS forwards to the real API using those, you never pass a
   raw upstream API key through this SDK.

```python
import agentraas

client = agentraas.Client(
    agentraas_key="ar_live_...",
    org_id="acme-corp",
    agent_id="billing-bot",
    base_url="http://localhost:13000",  # or your Cloud/self-hosted URL
)

result = client.call("stripe", "charge.create", {"amount": 5000, "currency": "usd"})
```

Or with dot-notation sugar:

```python
stripe = client.service("stripe")
result = stripe.charge.create({"amount": 5000, "currency": "usd"})
```

Calling a [Custom Action](https://github.com/sumedhchatse/agentraas#supported-services)
you've registered:

```python
result = client.custom("my-internal-webhook", {"foo": "bar"})
```

## Why `org_id` / `agent_id` matter

Omit them and every untagged SDK caller shares one unenforced identity
server-side — your per-agent rate limit and audit trail won't tell your
traffic apart from anyone else's. Set them once you've connected an
agent from the dashboard; it takes two extra kwargs.

## Retries are safe

AgentRaaS claims an atomic dedup slot server-side *before* forwarding
anything. If `client.call(...)` raises because of a timeout or a dropped
connection, calling it again with the same payload is safe — it either
completes normally or returns the cached result from the call that
actually went through. This SDK doesn't retry automatically; your own
retry logic (or your agent framework's) can be as aggressive as you want.

## Error handling

```python
from agentraas import AgentRaaSError

try:
    client.call("stripe", "charge.create", {"amount": 5000, "currency": "usd"})
except AgentRaaSError as err:
    print(err.status_code, err.req_id, str(err))
```

## License

MIT — see [LICENSE](./LICENSE). (The AgentRaaS server itself is
open-core; see the [main repo](https://github.com/sumedhchatse/agentraas)
for its licensing.)
