BREAKING: namespace change to codex_app_server_sdk

  - move implementation to canonical package namespace
  - remove codex_app_server_client compatibility shim package
  - update tests to import/patch codex_app_server_sdk modules directly
  - update packaging to ship only src/codex_app_server_sdk
  - clean README to remove legacy import migration block
  - bump version to 0.3.0

  BREAKING CHANGE: legacy imports from codex_app_server_client are no longer supported; use codex_app_server_sdk.

Author: emsi

README.md

@@ -1,4 +1,4 @@
-# codex-app-server-client
+# codex-app-server-sdk
 
 High-level async Python client for `codex app-server`.
 
@@ -47,7 +47,7 @@ uv pip install codex-app-server-sdk
 
 ```python
 import asyncio
-from codex_app_server_client import CodexClient
+from codex_app_server_sdk import CodexClient
 
 
 async def main() -> None:
@@ -70,7 +70,7 @@ You can override via:
 
 ```python
 import asyncio
-from codex_app_server_client import CodexClient
+from codex_app_server_sdk import CodexClient
 
 
 async def main() -> None:
@@ -92,7 +92,7 @@ Both high-level APIs support resuming the same running turn.
 
 ```python
 import asyncio
-from codex_app_server_client import CodexClient, CodexTurnInactiveError
+from codex_app_server_sdk import CodexClient, CodexTurnInactiveError
 
 
 async def main() -> None:
@@ -128,7 +128,7 @@ Use explicit thread handles when you need thread-scoped configuration.
 
 ```python
 import asyncio
-from codex_app_server_client import CodexClient, ThreadConfig, TurnOverrides
+from codex_app_server_sdk import CodexClient, ThreadConfig, TurnOverrides
 
 
 async def main() -> None:
@@ -176,7 +176,7 @@ asyncio.run(main())
 Example:
 
 ```python
-from codex_app_server_client import ThreadConfig, UNSET
+from codex_app_server_sdk import ThreadConfig, UNSET
 
 cfg = ThreadConfig(
     model=UNSET,  # omit key
@@ -376,7 +376,7 @@ uv run python examples/chat_session_websocket.py
 
 ## API reference (quick)
 
-### `CodexClient` (`src/codex_app_server_client/client.py`)
+### `CodexClient` (`src/codex_app_server_sdk/client.py`)
 
 - `connect_stdio(...)`: create a stdio-configured client (unstarted).
 - `connect_websocket(...)`: create a websocket-configured client (unstarted).
@@ -404,13 +404,13 @@ uv run python examples/chat_session_websocket.py
 - `interrupt_turn(turn_id, timeout=None)`: low-level turn interruption request.
 - `close()`: cancel receive loop and close transport.
 
-### `Transport` and implementations (`src/codex_app_server_client/transport.py`)
+### `Transport` and implementations (`src/codex_app_server_sdk/transport.py`)
 
 - `Transport.connect/send/recv/close`: abstract interface.
 - `StdioTransport`: line-delimited JSON over subprocess stdin/stdout.
 - `WebSocketTransport`: JSON messages over websocket frames.
 
-### Data models (`src/codex_app_server_client/models.py`)
+### Data models (`src/codex_app_server_sdk/models.py`)
 
 - `InitializeResult`: parsed initialize response (`protocol_version`, `server_info`, `capabilities`, `raw`).
 - `ConversationStep`: completed step from `chat(...)` (`step_type`, `item_type`, `text`, `item_id`, `thread_id`, `turn_id`, `data`).
@@ -422,7 +422,7 @@ uv run python examples/chat_session_websocket.py
 - `UNSET`: sentinel for “omit this field from request payload.”
 - `ApprovalPolicy`: literal type for approval policy values (`untrusted`, `on-failure`, `on-request`, `never`).
 
-### `ThreadHandle` (`src/codex_app_server_client/client.py`)
+### `ThreadHandle` (`src/codex_app_server_sdk/client.py`)
 
 - `thread_id`: bound thread id.
 - `defaults`: local thread config snapshot.
@@ -434,7 +434,7 @@ uv run python examples/chat_session_websocket.py
 - `set_name(name)`, `archive()`, `unarchive()`, `rollback(num_turns)`, `compact()`: thread lifecycle/history helpers.
 - `start_review(target, delivery=None)`: thread-bound review API.
 
-### Exceptions (`src/codex_app_server_client/errors.py`)
+### Exceptions (`src/codex_app_server_sdk/errors.py`)
 
 - `CodexError`: base exception.
 - `CodexTransportError`: transport/connectivity problems.
@@ -475,7 +475,7 @@ When `params=None`, the client sends:
 {
   "protocolVersion": "1",
   "clientInfo": {
-    "name": "codex-app-server-client",
+    "name": "codex-app-server-sdk",
     "version": "0.1.0"
   },
   "capabilities": {
@@ -520,7 +520,7 @@ Merge rules:
 
 ```python
 import asyncio
-from codex_app_server_client import CodexClient
+from codex_app_server_sdk import CodexClient
 
 
 async def main() -> None:

docs/api/client.md

@@ -1,3 +1,3 @@
-# `codex_app_server_client.client`
+# `codex_app_server_sdk.client`
 
-::: codex_app_server_client.client
+::: codex_app_server_sdk.client

docs/api/errors.md

@@ -1,3 +1,3 @@
-# `codex_app_server_client.errors`
+# `codex_app_server_sdk.errors`
 
-::: codex_app_server_client.errors
+::: codex_app_server_sdk.errors

docs/api/models.md

@@ -1,3 +1,3 @@
-# `codex_app_server_client.models`
+# `codex_app_server_sdk.models`
 
-::: codex_app_server_client.models
+::: codex_app_server_sdk.models

docs/api/package.md

@@ -1,3 +1,3 @@
-# `codex_app_server_client`
+# `codex_app_server_sdk`
 
-::: codex_app_server_client
+::: codex_app_server_sdk

docs/api/protocol.md

@@ -1,3 +1,3 @@
-# `codex_app_server_client.protocol`
+# `codex_app_server_sdk.protocol`
 
-::: codex_app_server_client.protocol
+::: codex_app_server_sdk.protocol

docs/api/transport.md

@@ -1,3 +1,3 @@
-# `codex_app_server_client.transport`
+# `codex_app_server_sdk.transport`
 
-::: codex_app_server_client.transport
+::: codex_app_server_sdk.transport

docs/behavior-guarantees.md

@@ -4,21 +4,21 @@ This page records intended high-level behavior for consumers.
 
 ## Related API
 
-- [`CodexClient.chat(...)`](api/client.md#codex_app_server_client.client.CodexClient.chat)
-- [`CodexClient.chat_once(...)`](api/client.md#codex_app_server_client.client.CodexClient.chat_once)
-- [`CodexTurnInactiveError`](api/errors.md#codex_app_server_client.errors.CodexTurnInactiveError)
-- [`CodexClient.cancel(...)`](api/client.md#codex_app_server_client.client.CodexClient.cancel)
-- [`CodexClient`](api/client.md#codex_app_server_client.client.CodexClient)
+- [`CodexClient.chat(...)`](api/client.md#codex_app_server_sdk.client.CodexClient.chat)
+- [`CodexClient.chat_once(...)`](api/client.md#codex_app_server_sdk.client.CodexClient.chat_once)
+- [`CodexTurnInactiveError`](api/errors.md#codex_app_server_sdk.errors.CodexTurnInactiveError)
+- [`CodexClient.cancel(...)`](api/client.md#codex_app_server_sdk.client.CodexClient.cancel)
+- [`CodexClient`](api/client.md#codex_app_server_sdk.client.CodexClient)
 
 ## Streaming model
 
-- [`chat(...)`](api/client.md#codex_app_server_client.client.CodexClient.chat) yields completed, non-delta step blocks.
+- [`chat(...)`](api/client.md#codex_app_server_sdk.client.CodexClient.chat) yields completed, non-delta step blocks.
 - step stream is based on live turn notifications (`item/completed`).
-- snapshot backfill from `thread/read` is not merged into [`chat(...)`](api/client.md#codex_app_server_client.client.CodexClient.chat).
+- snapshot backfill from `thread/read` is not merged into [`chat(...)`](api/client.md#codex_app_server_sdk.client.CodexClient.chat).
 
 ## Final text resolution
 
-- [`chat_once(...)`](api/client.md#codex_app_server_client.client.CodexClient.chat_once) prefers completed assistant messages from live events.
+- [`chat_once(...)`](api/client.md#codex_app_server_sdk.client.CodexClient.chat_once) prefers completed assistant messages from live events.
 - fallback to `thread/read(includeTurns=true)` is used when needed to recover final text.
 
 ## Timeout model
@@ -29,13 +29,13 @@ This page records intended high-level behavior for consumers.
 
 ## Continuation model
 
-- inactivity timeout raises [`CodexTurnInactiveError`](api/errors.md#codex_app_server_client.errors.CodexTurnInactiveError) with continuation token.
+- inactivity timeout raises [`CodexTurnInactiveError`](api/errors.md#codex_app_server_sdk.errors.CodexTurnInactiveError) with continuation token.
 - continuation can resume the same running turn in the same client instance.
 - continuation cannot be combined with fresh turn input/options.
 
 ## Cancel model
 
-- [`cancel(...)`](api/client.md#codex_app_server_client.client.CodexClient.cancel) sends best-effort interrupt.
+- [`cancel(...)`](api/client.md#codex_app_server_sdk.client.CodexClient.cancel) sends best-effort interrupt.
 - unread steps/events since continuation cursor are returned.
 - internal state is cleaned so thread reuse is safe.
 

docs/conversation.md

@@ -2,16 +2,16 @@
 
 ## Related API
 
-- [`CodexClient.chat_once(...)`](api/client.md#codex_app_server_client.client.CodexClient.chat_once)
-- [`CodexClient.chat(...)`](api/client.md#codex_app_server_client.client.CodexClient.chat)
-- [`CodexClient.start_thread(...)`](api/client.md#codex_app_server_client.client.CodexClient.start_thread)
-- [`ThreadHandle`](api/client.md#codex_app_server_client.client.ThreadHandle)
-- [`TurnOverrides`](api/models.md#codex_app_server_client.models.TurnOverrides)
+- [`CodexClient.chat_once(...)`](api/client.md#codex_app_server_sdk.client.CodexClient.chat_once)
+- [`CodexClient.chat(...)`](api/client.md#codex_app_server_sdk.client.CodexClient.chat)
+- [`CodexClient.start_thread(...)`](api/client.md#codex_app_server_sdk.client.CodexClient.start_thread)
+- [`ThreadHandle`](api/client.md#codex_app_server_sdk.client.ThreadHandle)
+- [`TurnOverrides`](api/models.md#codex_app_server_sdk.models.TurnOverrides)
 
 Two high-level conversation entrypoints are provided:
 
-- [`chat_once(...)`](api/client.md#codex_app_server_client.client.CodexClient.chat_once): one message in, final assistant text out
-- [`chat(...)`](api/client.md#codex_app_server_client.client.CodexClient.chat): async iterator of completed, non-delta step blocks
+- [`chat_once(...)`](api/client.md#codex_app_server_sdk.client.CodexClient.chat_once): one message in, final assistant text out
+- [`chat(...)`](api/client.md#codex_app_server_sdk.client.CodexClient.chat): async iterator of completed, non-delta step blocks
 
 ## `chat_once(...)`
 
@@ -34,12 +34,12 @@ async for step in client.chat("Diagnose this failure"):
 
 ### Step source semantics
 
-[`chat(...)`](api/client.md#codex_app_server_client.client.CodexClient.chat) emits steps from live `item/completed` notifications for the active
+[`chat(...)`](api/client.md#codex_app_server_sdk.client.CodexClient.chat) emits steps from live `item/completed` notifications for the active
 turn. It does not merge `thread/read` snapshot items into the same stream.
 
 ## Thread binding
 
-You can call APIs directly with `thread_id=...`, or use [`ThreadHandle`](api/client.md#codex_app_server_client.client.ThreadHandle):
+You can call APIs directly with `thread_id=...`, or use [`ThreadHandle`](api/client.md#codex_app_server_sdk.client.ThreadHandle):
 
 ```python
 thread = await client.start_thread()
@@ -49,6 +49,6 @@ result = await thread.chat_once("Hello")
 ## Per-turn metadata and overrides
 
 - `metadata` is applied on `turn/start`
-- [`TurnOverrides`](api/models.md#codex_app_server_client.models.TurnOverrides) controls per-turn execution options (`cwd`, `model`, `effort`, etc.)
+- [`TurnOverrides`](api/models.md#codex_app_server_sdk.models.TurnOverrides) controls per-turn execution options (`cwd`, `model`, `effort`, etc.)
 - when resuming with `continuation=...`, do not pass `text`, `thread_id`,
   `user`, `metadata`, `thread_config`, or `turn_overrides`

docs/examples.md

@@ -4,13 +4,13 @@ All examples live in `examples/`.
 
 ## Related API
 
-- [`CodexClient`](api/client.md#codex_app_server_client.client.CodexClient)
-- [`ThreadHandle`](api/client.md#codex_app_server_client.client.ThreadHandle)
-- [`CodexClient.chat(...)`](api/client.md#codex_app_server_client.client.CodexClient.chat)
-- [`CodexClient.chat_once(...)`](api/client.md#codex_app_server_client.client.CodexClient.chat_once)
-- [`CodexClient.cancel(...)`](api/client.md#codex_app_server_client.client.CodexClient.cancel)
-- [`CodexClient.start_thread(...)`](api/client.md#codex_app_server_client.client.CodexClient.start_thread)
-- [`CodexClient.resume_thread(...)`](api/client.md#codex_app_server_client.client.CodexClient.resume_thread)
+- [`CodexClient`](api/client.md#codex_app_server_sdk.client.CodexClient)
+- [`ThreadHandle`](api/client.md#codex_app_server_sdk.client.ThreadHandle)
+- [`CodexClient.chat(...)`](api/client.md#codex_app_server_sdk.client.CodexClient.chat)
+- [`CodexClient.chat_once(...)`](api/client.md#codex_app_server_sdk.client.CodexClient.chat_once)
+- [`CodexClient.cancel(...)`](api/client.md#codex_app_server_sdk.client.CodexClient.cancel)
+- [`CodexClient.start_thread(...)`](api/client.md#codex_app_server_sdk.client.CodexClient.start_thread)
+- [`CodexClient.resume_thread(...)`](api/client.md#codex_app_server_sdk.client.CodexClient.resume_thread)
 
 ## `chat_steps_rich.py`
 
@@ -43,7 +43,7 @@ uv run python examples/thread_resume_by_id.py --thread-id <existing-thread-id>
 
 ## `thread_concurrent_handles.py`
 
-Runs two fresh [`ThreadHandle`](api/client.md#codex_app_server_client.client.ThreadHandle)s concurrently over one shared client connection.
+Runs two fresh [`ThreadHandle`](api/client.md#codex_app_server_sdk.client.ThreadHandle)s concurrently over one shared client connection.
 
 ```bash
 uv run python examples/thread_concurrent_handles.py --transport stdio