tokenbee-sdk 2.0.0__tar.gz → 2.0.1__tar.gz

{tokenbee_sdk-2.0.0 → tokenbee_sdk-2.0.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: tokenbee-sdk
-Version: 2.0.0
+Version: 2.0.1
 Summary: Official Python SDK for TokenBee LLM inference gateway and observability.
 Author: TokenBee Inc.
 Author-email: "TokenBee Inc." <founders@tokenbee.io>
@@ -18,13 +18,13 @@ Dynamic: requires-python
 
 # TokenBee Python SDK
 
-Official Python SDK for [TokenBee](https://tokenbee.io) - The Intelligent LLM Inference Gateway with Observability, Compression, and Privacy.
+Official Python SDK for [TokenBee](https://tokenbee.io) — AI interaction capture, audit, replay, and optimization.
 
 ## Features
 
 - **Unified API**: Access multiple LLM providers (OpenAI, Anthropic, Google, Mistral, etc.) through a single interface.
 - **Intelligent Compression**: Reduce token usage and latency with context-aware compression.
-- **Privacy Guard**: Automatic PII masking and privacy-preserving inference.
+- **Configurable Capture**: Retain interaction content when you need it; turn capture off per request or globally.
 - **Built-in Observability**: Automatic tracking of latency, costs, and token usage.
 
 ## Links
@@ -82,7 +82,7 @@ response = client.send(
         "messages": [...],
         "compression": "auto",      # "auto" (default), "on", or "off"
         "rate": CompressionRate.HIGH, # MEDIUM (0.5), HIGH (0.33), etc.
-        "privacy": True
+        "capture": False            # per-request: do not retain message/response bodies
     }
 )
 ```
@@ -91,7 +91,7 @@ response = client.send(
 - **`rate`**: Controls the aggressiveness of compression. `HIGH` aims for ~67% token reduction.
 - **`sessionId`**: (Optional) String ID to group multiple requests into a single replayable session in the dashboard.
 - **`userId`**: (Optional) String ID to track usage and costs per unique end-user.
-- **`privacy`**: (Optional) Set to `True` to disable payload logging and session replays for this request. Metadata (latency, tokens) will still be recorded for observability.
+- **`capture`**: (Optional) Per-request override. `False` does not retain messages/responses (metadata like tokens/cost/latency can still be recorded). Precedence: per-request → project setting → default on.
 
 ### Supported Models
 

{tokenbee_sdk-2.0.0 → tokenbee_sdk-2.0.1}/README.md

@@ -1,12 +1,12 @@
 # TokenBee Python SDK
 
-Official Python SDK for [TokenBee](https://tokenbee.io) - The Intelligent LLM Inference Gateway with Observability, Compression, and Privacy.
+Official Python SDK for [TokenBee](https://tokenbee.io) — AI interaction capture, audit, replay, and optimization.
 
 ## Features
 
 - **Unified API**: Access multiple LLM providers (OpenAI, Anthropic, Google, Mistral, etc.) through a single interface.
 - **Intelligent Compression**: Reduce token usage and latency with context-aware compression.
-- **Privacy Guard**: Automatic PII masking and privacy-preserving inference.
+- **Configurable Capture**: Retain interaction content when you need it; turn capture off per request or globally.
 - **Built-in Observability**: Automatic tracking of latency, costs, and token usage.
 
 ## Links
@@ -64,7 +64,7 @@ response = client.send(
         "messages": [...],
         "compression": "auto",      # "auto" (default), "on", or "off"
         "rate": CompressionRate.HIGH, # MEDIUM (0.5), HIGH (0.33), etc.
-        "privacy": True
+        "capture": False            # per-request: do not retain message/response bodies
     }
 )
 ```
@@ -73,7 +73,7 @@ response = client.send(
 - **`rate`**: Controls the aggressiveness of compression. `HIGH` aims for ~67% token reduction.
 - **`sessionId`**: (Optional) String ID to group multiple requests into a single replayable session in the dashboard.
 - **`userId`**: (Optional) String ID to track usage and costs per unique end-user.
-- **`privacy`**: (Optional) Set to `True` to disable payload logging and session replays for this request. Metadata (latency, tokens) will still be recorded for observability.
+- **`capture`**: (Optional) Per-request override. `False` does not retain messages/responses (metadata like tokens/cost/latency can still be recorded). Precedence: per-request → project setting → default on.
 
 ### Supported Models
 

{tokenbee_sdk-2.0.0 → tokenbee_sdk-2.0.1}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "tokenbee-sdk"
-version = "2.0.0"
+version = "2.0.1"
 authors = [
   { name="TokenBee Inc.", email="founders@tokenbee.io" },
 ]

{tokenbee_sdk-2.0.0 → tokenbee_sdk-2.0.1}/setup.py

@@ -6,7 +6,7 @@ long_description = (this_directory / "README.md").read_text()
 
 setup(
     name="tokenbee-sdk",
-    version="2.0.0",
+    version="2.0.1",
     packages=find_packages(),
     install_requires=["httpx>=0.23.0"],
     author="TokenBee Inc.",

{tokenbee_sdk-2.0.0 → tokenbee_sdk-2.0.1}/tokenbee/__init__.py

@@ -1,6 +1,7 @@
 import os
 import httpx
 from enum import Enum
+from typing import Optional
 
 class CompressionRate(str, Enum):
     LOW = "0.75"
@@ -90,7 +91,7 @@ class TokenBee:
         context: str = TokenBeeContext.AUTO,
         model: str = "", 
         provider: str = "",
-        privacy: bool = False
+        capture: Optional[bool] = None
     ):
         self.api_key = api_key
         self.llm_key = llm_key
@@ -104,8 +105,9 @@ class TokenBee:
             "X-TokenBee-Rate": rate,
             "X-TokenBee-Strategy": strategy,
             "X-TokenBee-Context": context,
-            "X-TokenBee-Privacy": str(privacy).lower()
         }
+        if capture is not None:
+            self.headers["X-TokenBee-Capture"] = str(capture).lower()
         if model:
             self.headers["X-TokenBee-Model"] = model
         if provider:
@@ -130,8 +132,8 @@ class TokenBee:
             headers["X-TokenBee-Strategy"] = str(input["strategy"])
         if "context" in input:
             headers["X-TokenBee-Context"] = str(input["context"])
-        if "privacy" in input:
-            headers["X-TokenBee-Privacy"] = str(input["privacy"]).lower()
+        if "capture" in input:
+            headers["X-TokenBee-Capture"] = str(input["capture"]).lower()
         if "sessionId" in input:
             headers["X-TB-Session-Id"] = str(input["sessionId"])
         if "userId" in input:
@@ -143,7 +145,7 @@ class TokenBee:
         payload.pop("rate", None)
         payload.pop("strategy", None)
         payload.pop("context", None)
-        payload.pop("privacy", None)
+        payload.pop("capture", None)
         payload.pop("sessionId", None)
         payload.pop("userId", None)
 

{tokenbee_sdk-2.0.0 → tokenbee_sdk-2.0.1}/tokenbee_sdk.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: tokenbee-sdk
-Version: 2.0.0
+Version: 2.0.1
 Summary: Official Python SDK for TokenBee LLM inference gateway and observability.
 Author: TokenBee Inc.
 Author-email: "TokenBee Inc." <founders@tokenbee.io>
@@ -18,13 +18,13 @@ Dynamic: requires-python
 
 # TokenBee Python SDK
 
-Official Python SDK for [TokenBee](https://tokenbee.io) - The Intelligent LLM Inference Gateway with Observability, Compression, and Privacy.
+Official Python SDK for [TokenBee](https://tokenbee.io) — AI interaction capture, audit, replay, and optimization.
 
 ## Features
 
 - **Unified API**: Access multiple LLM providers (OpenAI, Anthropic, Google, Mistral, etc.) through a single interface.
 - **Intelligent Compression**: Reduce token usage and latency with context-aware compression.
-- **Privacy Guard**: Automatic PII masking and privacy-preserving inference.
+- **Configurable Capture**: Retain interaction content when you need it; turn capture off per request or globally.
 - **Built-in Observability**: Automatic tracking of latency, costs, and token usage.
 
 ## Links
@@ -82,7 +82,7 @@ response = client.send(
         "messages": [...],
         "compression": "auto",      # "auto" (default), "on", or "off"
         "rate": CompressionRate.HIGH, # MEDIUM (0.5), HIGH (0.33), etc.
-        "privacy": True
+        "capture": False            # per-request: do not retain message/response bodies
     }
 )
 ```
@@ -91,7 +91,7 @@ response = client.send(
 - **`rate`**: Controls the aggressiveness of compression. `HIGH` aims for ~67% token reduction.
 - **`sessionId`**: (Optional) String ID to group multiple requests into a single replayable session in the dashboard.
 - **`userId`**: (Optional) String ID to track usage and costs per unique end-user.
-- **`privacy`**: (Optional) Set to `True` to disable payload logging and session replays for this request. Metadata (latency, tokens) will still be recorded for observability.
+- **`capture`**: (Optional) Per-request override. `False` does not retain messages/responses (metadata like tokens/cost/latency can still be recorded). Precedence: per-request → project setting → default on.
 
 ### Supported Models