Merge current subagent naming stack into Team session UI

python/development.i18n.yaml

@@ -9,8 +9,8 @@
   en: 8e99640e4936332d
   zh: 729d5de1e0ad2869
 /python-contributor-workflows/validate-the-sdk:
-  en: 37884fe76e167a0e
-  zh: 4b6e50768c4bb945
+  en: ad4d61dccf2756a4
+  zh: 9365affc91287738
 /python-contributor-workflows/run-against-node-source:
   en: 940c941d189a8248
   zh: 321f5183383d4a12

python/development.md

@@ -38,7 +38,7 @@ uv run --project python/sdk python scripts/smoke-python-runtime.py \
   --scenario sdk-minimal --exe dist-exe/deepseek-harness-sdk-runtime-macos-arm64
 ```
 
-SDK scenarios compare committed expected output under `scripts/snapshots/python-sdk-single-exe/`. `minimal/model-visible.json` pins the Linux/macOS `sdk-minimal` profile's assembled system prompts, advertised tool schemas, and model-visible messages; `minimal/win-x64/model-visible.json` pins its PowerShell counterpart. `authoring/model-visible.json` pins the Linux/macOS Office authoring catalog and dependency result; `authoring/win-x64/model-visible.json` pins the Windows catalog, including the ACL diagnostic skill. A plugin that contributes an unintended system section or user message therefore fails the job, and every message the profile emits is compared. `advanced/` pins one complex process's SDK result and parent/child session logs across every target. `restart/` launches two complete SDK runtime processes against one persistence root and snapshots their isolated model histories, high-level results, and separate durable logs across every target. `sdk-minimal-in-history` reuses the persistent-shell and editor scenario with a section that changes after the first successful shell call. `minimal-in-history/prompt-history.json` pins both prompt versions, the unchanged leading prompt in later requests, appended SDK system-message events, and `request/context.systemPromptUpdate`; tool schemas stay fixed and the editor file is checked independently. `sdk-dynamic-tools` pins native tool additions and removals in `dynamic-tools/tool-history.json`, including historical header references and agreement between Python SDK results, notifications, and persisted events. Rerun the owning scenario with `--update-snapshots` and review that diff before committing it.
+SDK scenarios compare committed expected output under `scripts/snapshots/python-sdk-single-exe/`. `minimal/model-visible.json` pins the Linux/macOS `sdk-minimal` profile's assembled system prompts, advertised tool schemas, and model-visible messages; `minimal/win-x64/model-visible.json` pins its PowerShell counterpart. `authoring/model-visible.json` pins the Linux/macOS Office authoring catalog and dependency result; `authoring/win-x64/model-visible.json` pins the Windows catalog, including the ACL diagnostic skill. A plugin that contributes an unintended system section or user message therefore fails the job, and every message the profile emits is compared. `advanced/` pins one complex process's SDK result and parent/child session logs across every target. `restart/` launches two complete SDK runtime processes against one persistence root and snapshots their isolated model histories, high-level results, and separate durable logs across every target. Its resumed turn initializes a different model and pins the persisted selection, request route, and SDK event sequence. `sdk-minimal-in-history` reuses the persistent-shell and editor scenario with a section that changes after the first successful shell call. `minimal-in-history/prompt-history.json` pins both prompt versions, the unchanged leading prompt in later requests, appended SDK system-message events, and `request/context.systemPromptUpdate`; tool schemas stay fixed and the editor file is checked independently. `sdk-dynamic-tools` pins native tool additions and removals in `dynamic-tools/tool-history.json`, including historical header references and agreement between Python SDK results, notifications, and persisted events. Rerun the owning scenario with `--update-snapshots` and review that diff before committing it.
 
 `scheduler-recovery/` records a failed tool turn and a subsequent completed turn. It preserves the original `UNKNOWN` turn error, the first tool's completed result, and `TOOL_OUTCOME_UNKNOWN` / `TOOL_NOT_STARTED` results for the remaining calls. `sdk-snapshot` and `all` include this case; `--scenario sdk-recovery` runs it alone.
 

python/development.zh.md

@@ -38,7 +38,7 @@ uv run --project python/sdk python scripts/smoke-python-runtime.py \
   --scenario sdk-minimal --exe dist-exe/deepseek-harness-sdk-runtime-macos-arm64
 ```
 
-SDK 场景会比对 `scripts/snapshots/python-sdk-single-exe/` 下已提交的期望输出。`minimal/model-visible.json` 固定 Linux／macOS `sdk-minimal` profile 所组装的系统提示词、对外公布的工具 schema 与模型可见消息；`minimal/win-x64/model-visible.json` 固定对应的 PowerShell 版本。`authoring/model-visible.json` 固定 Linux／macOS 的 Office 创作技能目录和依赖结果；`authoring/win-x64/model-visible.json` 固定 Windows 的技能目录，其中包含 ACL 诊断技能。因此，插件一旦贡献出计划外的系统分段或 user 消息，该任务即失败，且该 profile 发出的每条消息都会参与比对。`advanced/` 跨所有目标固定一个复杂进程的 SDK 结果及父／子会话日志。`restart/` 针对同一持久化根目录启动两个完整 SDK 运行时进程，并跨所有目标固定其彼此隔离的模型历史、高层结果与独立持久日志。`sdk-minimal-in-history` 复用持久 shell 与编辑器场景，并在首次 shell 调用成功后更改一个分段。`minimal-in-history/prompt-history.json` 固定两个提示词版本、后续请求中不变的首条提示词、追加的 SDK system-message 事件及 `request/context.systemPromptUpdate`；工具 schema 保持不变，并独立检查编辑器创建的文件。`sdk-dynamic-tools` 在 `dynamic-tools/tool-history.json` 中固定原生工具的新增与移除，包括历史 header 引用，以及 Python SDK 结果、通知与持久事件的一致性。重新运行对应场景时加上 `--update-snapshots`，并在提交前审阅该差异。
+SDK 场景会比对 `scripts/snapshots/python-sdk-single-exe/` 下已提交的期望输出。`minimal/model-visible.json` 固定 Linux／macOS `sdk-minimal` profile 所组装的系统提示词、对外公布的工具 schema 与模型可见消息；`minimal/win-x64/model-visible.json` 固定对应的 PowerShell 版本。`authoring/model-visible.json` 固定 Linux／macOS 的 Office 创作技能目录和依赖结果；`authoring/win-x64/model-visible.json` 固定 Windows 的技能目录，其中包含 ACL 诊断技能。因此，插件一旦贡献出计划外的系统分段或 user 消息，该任务即失败，且该 profile 发出的每条消息都会参与比对。`advanced/` 跨所有目标固定一个复杂进程的 SDK 结果及父／子会话日志。`restart/` 针对同一持久化根目录启动两个完整 SDK 运行时进程，并跨所有目标固定其彼此隔离的模型历史、高层结果与独立持久日志。恢复后的轮次初始化另一个模型，并固定持久模型选择、请求路由与 SDK 事件顺序。`sdk-minimal-in-history` 复用持久 shell 与编辑器场景，并在首次 shell 调用成功后更改一个分段。`minimal-in-history/prompt-history.json` 固定两个提示词版本、后续请求中不变的首条提示词、追加的 SDK system-message 事件及 `request/context.systemPromptUpdate`；工具 schema 保持不变，并独立检查编辑器创建的文件。`sdk-dynamic-tools` 在 `dynamic-tools/tool-history.json` 中固定原生工具的新增与移除，包括历史 header 引用，以及 Python SDK 结果、通知与持久事件的一致性。重新运行对应场景时加上 `--update-snapshots`，并在提交前审阅该差异。
 
 `scheduler-recovery/` 记录一个失败的工具轮次及后续完成的轮次。它保留原始的 `UNKNOWN` 轮次错误、第一个工具已完成的结果，以及其余调用的 `TOOL_OUTCOME_UNKNOWN` / `TOOL_NOT_STARTED` 结果。`sdk-snapshot` 和 `all` 包含此场景；`--scenario sdk-recovery` 单独运行它。
 

python/sdk-runtime/package.json

@@ -14,7 +14,6 @@
     "@deepseek-ai/dsh": "workspace:*",
     "@deepseek-ai/dsh-acp": "workspace:*",
     "@deepseek-ai/dsh-agent": "workspace:*",
-    "@deepseek-ai/dsh-agent-controller": "workspace:*",
     "@deepseek-ai/dsh-agent-default-model": "workspace:*",
     "@deepseek-ai/dsh-agent-instructions": "workspace:*",
     "@deepseek-ai/dsh-agent-loop": "workspace:*",
@@ -111,6 +110,7 @@
     "@deepseek-ai/dsh-storage": "workspace:*",
     "@deepseek-ai/dsh-storage-domain": "workspace:*",
     "@deepseek-ai/dsh-subagent": "workspace:*",
+    "@deepseek-ai/dsh-subagent-names": "workspace:*",
     "@deepseek-ai/dsh-subagent-acp": "workspace:*",
     "@deepseek-ai/dsh-subagent-fork-in-process": "workspace:*",
     "@deepseek-ai/dsh-subagent-spawn-in-process": "workspace:*",
@@ -128,6 +128,7 @@
     "@deepseek-ai/dsh-tool-bash-persistent": "workspace:*",
     "@deepseek-ai/dsh-tool-call-timeout-policy": "workspace:*",
     "@deepseek-ai/dsh-tool-cordis": "workspace:*",
+    "@deepseek-ai/dsh-tool-executor-ptc": "workspace:*",
     "@deepseek-ai/dsh-tool-fs": "workspace:*",
     "@deepseek-ai/dsh-tool-fs-search": "workspace:*",
     "@deepseek-ai/dsh-tool-goal": "workspace:*",

python/sdk/README.i18n.yaml

@@ -9,8 +9,11 @@
   en: ddf4dabea526cef1
   zh: 6e31cab10d2f3ed8
 /deepseek-harness-python-sdk/start-a-runtime:
-  en: 5fceaffcc35039c6
-  zh: 1d87e3defe7e875b
+  en: 1ae09bebf0f9e19b
+  zh: 55a0d557a2cae4d8
+/deepseek-harness-python-sdk/client-approvals:
+  en: fc69be1beb8abff1
+  zh: 91b56781931a6f68
 /deepseek-harness-python-sdk/customize-plugins:
   en: 69a6a6745c6e89c1
   zh: 85e82a4cc31e16e5

python/sdk/README.md

@@ -58,6 +58,27 @@ with DeepSeekHarness(dsh_home="/absolute/path/to/saved-home") as harness:
     result = session.run("Continue the task.")
 ```
 
+<a id="client-approvals"></a>
+## Client approvals
+
+This low-level example receives one approval and rejects it. Use a dedicated request-handling thread or loop for a full run: `next_request()` blocks, and the request handler must run concurrently with `session/wait` or a high-level `run()`. Set `approval_requests=True` on `HarnessClient.initialize()` or `DeepSeekHarnessConfig` only when that handler is ready. Omission preserves the runtime's answerer chain. Decisions are one-shot; invalid responses and lost connections fail closed, and cancellation discards late replies. `ApprovalRequestResult` validates response values when constructing decisions.
+
+```py
+from deepseek_harness import ApprovalRequestParams, HarnessClient, HarnessConfig
+
+with HarnessClient(HarnessConfig(dsh_home="/absolute/path/to/dsh-home")) as client:
+    client.initialize(cwd="/absolute/path/to/workspace", provider="deepseek-official",
+                      model="deepseek-v4-flash", approval_requests=True)
+    client.session_prompt("main", [{"type": "text", "text": "Run the operation requiring approval."}])
+    request = client.next_request()
+    if request.method == "approval/request":
+        question = ApprovalRequestParams.model_validate(request.payload)
+        print(question.toolName, question.reason)
+        client.respond(request.id, {"outcome": "rejected"})
+    else:
+        client.respond_error(request.id, code=-32601, message="Unsupported client request")
+```
+
 ## Customize plugins
 
 Persistent customization belongs to a `dsh` profile. Initialize the shipped SDK profile and install an external bundle with the runtime wheel's `dsh` command:

python/sdk/README.zh.md

@@ -58,6 +58,27 @@ with DeepSeekHarness(dsh_home="/absolute/path/to/saved-home") as harness:
     result = session.run("Continue the task.")
 ```
 
+<a id="client-approvals"></a>
+## 客户端审批
+
+此底层示例接收一次审批并拒绝。完整运行应使用专门处理请求的线程或循环：`next_request()` 会阻塞，请求处理器必须与 `session/wait` 或高层 `run()` 并行运行。仅在处理器就绪时，在 `HarnessClient.initialize()` 或 `DeepSeekHarnessConfig` 设置 `approval_requests=True`；省略时保留运行时应答链。决定只作用于单次操作；无效响应和断连会拒绝放行，取消后丢弃迟到响应。构造决定时，可用 `ApprovalRequestResult` 校验响应值。
+
+```py
+from deepseek_harness import ApprovalRequestParams, HarnessClient, HarnessConfig
+
+with HarnessClient(HarnessConfig(dsh_home="/absolute/path/to/dsh-home")) as client:
+    client.initialize(cwd="/absolute/path/to/workspace", provider="deepseek-official",
+                      model="deepseek-v4-flash", approval_requests=True)
+    client.session_prompt("main", [{"type": "text", "text": "Run the operation requiring approval."}])
+    request = client.next_request()
+    if request.method == "approval/request":
+        question = ApprovalRequestParams.model_validate(request.payload)
+        print(question.toolName, question.reason)
+        client.respond(request.id, {"outcome": "rejected"})
+    else:
+        client.respond_error(request.id, code=-32601, message="Unsupported client request")
+```
+
 ## 自定义插件
 
 持久自定义属于 `dsh` profile。使用运行时 wheel 包提供的 `dsh` 命令初始化随附的 SDK profile，并安装外部 bundle：

python/sdk/src/deepseek_harness/__init__.py

@@ -1,7 +1,10 @@
 from .api import DeepSeekHarness, DeepSeekHarnessConfig, RunResult, Session
 from .client import HarnessClient, HarnessConfig, UserQuestionHandler
 from .errors import SdkProtocolError
-from .models import IncomingRequest, InitializeResponse, JsonObject, Notification, ServerInfo, UserQuestionAnswer, UserQuestionAnswerItem, UserQuestionRequest
+from .models import (
+    ApprovalRequestParams, ApprovalRequestResult, IncomingRequest, InitializeResponse,
+    JsonObject, Notification, ServerInfo, UserQuestionAnswer, UserQuestionAnswerItem, UserQuestionRequest,
+)
 
 __all__ = [
     "DeepSeekHarness",
@@ -15,6 +18,8 @@ __all__ = [
     "UserQuestionAnswer",
     "UserQuestionAnswerItem",
     "SdkProtocolError",
+    "ApprovalRequestParams",
+    "ApprovalRequestResult",
     "IncomingRequest",
     "InitializeResponse",
     "JsonObject",

python/sdk/src/deepseek_harness/api.py

@@ -36,6 +36,7 @@ class DeepSeekHarnessConfig:
     base_url: str | None = None
     api_key: str | None = None
     on_user_question: UserQuestionHandler | None = None
+    approval_requests: bool = False
 
 
 @dataclass(slots=True)
@@ -112,6 +113,7 @@ class DeepSeekHarness:
             model=self.config.model,
             reasoning_effort=self.config.reasoning_effort,
             max_tokens=self.config.max_tokens,
+            approval_requests=self.config.approval_requests,
         )
         self._initialized = True
 

python/sdk/src/deepseek_harness/client.py

@@ -156,12 +156,21 @@ class HarnessClient:
         model: str,
         reasoning_effort: str | None = None,
         max_tokens: int | None = None,
+        approval_requests: bool = False,
     ) -> InitializeResponse:
+        """Initialize the runtime; approval_requests enables approval/request reception.
+
+        When enabled, drain next_request() concurrently with blocking waits and
+        return one-operation decisions through respond(). Omission preserves the
+        runtime's composed approval answerers.
+        """
         payload: JsonObject = {
             "cwd": str(Path(cwd).resolve()),
             "provider": provider,
             "model": model,
         }
+        if approval_requests:
+            payload["approvalRequests"] = True
         if reasoning_effort is not None:
             payload["reasoningEffort"] = reasoning_effort
         if max_tokens is not None:

python/sdk/src/deepseek_harness/models.py

@@ -1,7 +1,7 @@
 from __future__ import annotations
 
 from dataclasses import dataclass
-from typing import TypeAlias
+from typing import Literal, TypeAlias
 
 from pydantic import BaseModel, Field, JsonValue as PydanticJsonValue
 
@@ -32,6 +32,20 @@ class InitializeResponse(BaseModel):
     serverInfo: ServerInfo | None = None
 
 
+class ApprovalRequestParams(BaseModel):
+    """Server-to-client approval/request payload after explicit handshake opt-in."""
+
+    sessionId: str
+    toolName: str
+    callId: str | None = None
+    reason: str | None = None
+
+
+class ApprovalRequestResult(BaseModel):
+    """One-operation decision; allowed-once is the only grant."""
+
+    outcome: Literal["allowed-once", "rejected", "cancelled"]
+
 class UserQuestionRequest(BaseModel):
     """Runtime question batch; questions retain their options, detail, and presentation intent."""
 

python/sdk/tests/test_approval.py

@@ -0,0 +1,55 @@
+"""Approval request decoding and explicit wire capability negotiation."""
+from __future__ import annotations
+
+import sys
+from pathlib import Path
+
+import pytest
+from pydantic import ValidationError
+
+from deepseek_harness import ApprovalRequestParams, ApprovalRequestResult, DeepSeekHarnessConfig, HarnessClient
+
+
+@pytest.mark.parametrize("enabled", [False, True])
+def test_approval_handshake_and_response(tmp_path: Path, enabled: bool) -> None:
+    peer = tmp_path / "approval_peer.py"
+    peer.write_text('''import json
+import sys
+for line in sys.stdin:
+    msg = json.loads(line)
+    if msg.get("method") == "initialize":
+        print(json.dumps({"jsonrpc": "2.0", "id": msg["id"], "result": {"serverInfo": {"name": "test"}}}), flush=True)
+        print(json.dumps({"jsonrpc": "2.0", "method": "capability", "params": msg["params"]}), flush=True)
+        if msg["params"].get("approvalRequests"):
+            print(json.dumps({"jsonrpc": "2.0", "id": "ask", "method": "approval/request", "params": {"sessionId": "main", "toolName": "bash", "reason": "write"}}), flush=True)
+    elif msg.get("id") == "ask":
+        print(json.dumps({"jsonrpc": "2.0", "method": "decision", "params": msg["result"]}), flush=True)
+    elif msg.get("method") == "shutdown":
+        print(json.dumps({"jsonrpc": "2.0", "id": msg["id"], "result": {}}), flush=True)
+        break
+''')
+    with HarnessClient(_launch_args=(sys.executable, str(peer))) as client:
+        client.initialize(cwd=str(tmp_path), provider="mock", model="mock", approval_requests=enabled)
+        payload = client.next_notification().payload
+        assert payload.get("approvalRequests", False) is enabled
+        if not enabled:
+            assert "approvalRequests" not in payload
+            return
+        request = client.next_request()
+        assert request.method == "approval/request"
+        question = ApprovalRequestParams.model_validate(request.payload)
+        assert question.toolName == "bash"
+        assert question.callId is None
+        client.respond(request.id, ApprovalRequestResult(outcome="rejected").model_dump())
+        assert client.next_notification().payload == {"outcome": "rejected"}
+
+
+def test_approval_result_rejects_unrecognized_grants() -> None:
+    with pytest.raises(ValidationError):
+        ApprovalRequestResult(outcome="allow-always")
+
+
+def test_approval_opt_in_preserves_existing_config_positional_arguments() -> None:
+    config = DeepSeekHarnessConfig("mock", "model", None, None, "/workspace")
+    assert config.cwd == "/workspace"
+    assert config.approval_requests is False

python/sdk/tests/test_approval_runtime.py

@@ -0,0 +1,52 @@
+"""Python decisions control a guarded file operation through the built sdk profile."""
+from __future__ import annotations
+
+import json
+import shutil
+from concurrent.futures import ThreadPoolExecutor
+from pathlib import Path
+
+import pytest
+from pydantic import BaseModel
+
+from deepseek_harness import ApprovalRequestParams, HarnessClient, HarnessConfig
+
+
+class EmptyResponse(BaseModel):
+    pass
+
+
+@pytest.mark.parametrize("outcome", ["allowed-once", "rejected"])
+def test_python_approval_controls_runtime_operation(tmp_path: Path, outcome: str) -> None:
+    repo = Path(__file__).resolve().parents[3]
+    node = shutil.which("node")
+    binary = repo / "apps/cli/lib/bin.js"
+    if node is None or not binary.exists():
+        pytest.skip("requires Node and built dsh CLI")
+    fixture = repo / "packages/sdk/server/tests/fixtures/approval.mjs"
+    marker = tmp_path / "approved.txt"
+    patch = tmp_path / "approval.patch.yml"
+    patch.write_text("- insert:\n    - id: approval-fixture\n      name: " + json.dumps(str(fixture))
+                     + "\n      config:\n        stopAfterApproval: true\n        path: " + json.dumps(str(marker)) + "\n")
+    # The client closes inside the executor lifetime, unblocking a reader even
+    # when a regression prevents the expected request from arriving.
+    with ThreadPoolExecutor(max_workers=1) as worker:
+        with HarnessClient(
+            HarnessConfig(cwd=str(repo), request_timeout_seconds=20, shutdown_timeout_seconds=20,
+                          env={"DSH_HOME": str(tmp_path / "home"), "DSH_PERMISSION_MODE": "workspace-write",
+                               "DSH_TELEMETRY_DISABLED": "1", "DEEPSEEK_API_KEY": "keyless-no-model-call"}),
+            _launch_args=(node, str(binary), "--profile", "sdk", "--patch", str(patch)),
+        ) as client:
+            client.initialize(cwd=str(tmp_path), provider="deepseek-official", model="deepseek-v4-flash",
+                              approval_requests=True)
+            client.session_prompt("main", [{"type": "text", "text": "Request the marker write."}])
+            request = worker.submit(client.next_request).result(timeout=20)
+            assert request.method == "approval/request"
+            question = ApprovalRequestParams.model_validate(request.payload)
+            assert question.sessionId == "main"
+            assert question.toolName == "fixture-write"
+            client.respond(request.id, {"outcome": outcome})
+            client.request("session/wait", {"sessionId": "main"}, response_model=EmptyResponse)
+            assert marker.exists() is (outcome == "allowed-once")
+            if marker.exists():
+                assert marker.read_text() == "approved\n"

python/sdk/tests/test_client.py

@@ -165,6 +165,7 @@ for line in sys.stdin:
     )
 
     with DeepSeekHarness(
+        approval_requests=True,
         model="deepseek-v4-flash",
         reasoning_effort="max",
         max_tokens=4096,
@@ -211,6 +212,7 @@ for line in sys.stdin:
     assert dumped_env["DSH_SESSION_ROOT"] is None
     assert dumped_env["DSH_CORDIS_CONFIG"] is None
     assert json.loads(init_dump.read_text()) == {
+        "approvalRequests": True,
         "cwd": str(tmp_path),
         "provider": "deepseek-official",
         "model": "deepseek-v4-flash",