deepagents-code 0.1.3__tar.gz → 0.1.4__tar.gz

{deepagents_code-0.1.3 → deepagents_code-0.1.4}/CHANGELOG.md

@@ -1,5 +1,21 @@
 # Deep Agents Code Changelog
 
+## [0.1.4](https://github.com/langchain-ai/deepagents/compare/deepagents-code==0.1.3...deepagents-code==0.1.4) (2026-05-23)
+
+### Features
+
+* Add `--sandbox-snapshot-name` flag ([#3538](https://github.com/langchain-ai/deepagents/issues/3538)) ([b01392e](https://github.com/langchain-ai/deepagents/commit/b01392e7549798434f27f3784fa8c4e734053787))
+* `dcode mcp config` and unify `--mcp-config` flag ([#3541](https://github.com/langchain-ai/deepagents/issues/3541)) ([f037b14](https://github.com/langchain-ai/deepagents/commit/f037b140f90a1ba3725b3ef23ab385b3cafe223b))
+* Interpreter middleware via `langchain-quickjs` ([#3525](https://github.com/langchain-ai/deepagents/issues/3525)) ([f0ca89c](https://github.com/langchain-ai/deepagents/commit/f0ca89c962c22058194121526638bc2d29f546bd))
+
+### Bug Fixes
+
+* Chat input history navigation and newline scrolling ([#3560](https://github.com/langchain-ai/deepagents/issues/3560)) ([3b51cbd](https://github.com/langchain-ai/deepagents/commit/3b51cbdc8c50d9990477e18a47de6a58e9165bab))
+* Distinguish LangSmith failure modes in `/trace` ([#3558](https://github.com/langchain-ai/deepagents/issues/3558)) ([4d158a0](https://github.com/langchain-ai/deepagents/commit/4d158a031aecad8862e02e332f127573003938ec))
+* Recover initial session prompts from writes table ([#3535](https://github.com/langchain-ai/deepagents/issues/3535)) ([46b6f3f](https://github.com/langchain-ai/deepagents/commit/46b6f3f3e6ce880cd5ec9cf59622bb745d6ac2eb))
+* Install script binary checks reference `dcode` ([#3546](https://github.com/langchain-ai/deepagents/issues/3546)) ([f8977a6](https://github.com/langchain-ai/deepagents/commit/f8977a63769e3f2037619f32596cb9bb7bd1020b))
+* Show tool call previews during batched HITL approvals ([#3530](https://github.com/langchain-ai/deepagents/issues/3530)) ([84daa1a](https://github.com/langchain-ai/deepagents/commit/84daa1a2e27963a6d7694dc9278de83782b4a7b7))
+
 ## [0.1.3](https://github.com/langchain-ai/deepagents/compare/deepagents-code==0.1.2...deepagents-code==0.1.3) (2026-05-20)
 
 ### Features

{deepagents_code-0.1.3 → deepagents_code-0.1.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: deepagents-code
-Version: 0.1.3
+Version: 0.1.4
 Summary: Terminal interface for Deep Agents - interactive AI agent with file operations, shell access, and sub-agent capabilities.
 Project-URL: Homepage, https://docs.langchain.com/oss/python/deepagents/overview
 Project-URL: Documentation, https://reference.langchain.com/python/deepagents/
@@ -69,7 +69,7 @@ Requires-Dist: langchain-google-vertexai<4.0.0,>=3.2.3; extra == 'all-providers'
 Requires-Dist: langchain-groq<2.0.0,>=1.1.2; extra == 'all-providers'
 Requires-Dist: langchain-huggingface<2.0.0,>=1.2.2; extra == 'all-providers'
 Requires-Dist: langchain-ibm<2.0.0,>=1.0.7; extra == 'all-providers'
-Requires-Dist: langchain-litellm<2.0.0,>=0.6.4; extra == 'all-providers'
+Requires-Dist: langchain-litellm<2.0.0,>=0.6.6; extra == 'all-providers'
 Requires-Dist: langchain-mistralai<2.0.0,>=1.1.4; extra == 'all-providers'
 Requires-Dist: langchain-nvidia-ai-endpoints<2.0.0,>=1.2.1; extra == 'all-providers'
 Requires-Dist: langchain-ollama<2.0.0,>=1.1.0; extra == 'all-providers'
@@ -106,7 +106,7 @@ Requires-Dist: langchain-huggingface<2.0.0,>=1.2.2; extra == 'huggingface'
 Provides-Extra: ibm
 Requires-Dist: langchain-ibm<2.0.0,>=1.0.7; extra == 'ibm'
 Provides-Extra: litellm
-Requires-Dist: langchain-litellm<2.0.0,>=0.6.4; extra == 'litellm'
+Requires-Dist: langchain-litellm<2.0.0,>=0.6.6; extra == 'litellm'
 Provides-Extra: mistralai
 Requires-Dist: langchain-mistralai<2.0.0,>=1.1.4; extra == 'mistralai'
 Provides-Extra: modal
@@ -121,6 +121,8 @@ Provides-Extra: openrouter
 Requires-Dist: langchain-openrouter<2.0.0,>=0.2.3; extra == 'openrouter'
 Provides-Extra: perplexity
 Requires-Dist: langchain-perplexity<2.0.0,>=1.2.0; extra == 'perplexity'
+Provides-Extra: quickjs
+Requires-Dist: langchain-quickjs<0.2.0,>=0.1.2; extra == 'quickjs'
 Provides-Extra: runloop
 Requires-Dist: langchain-runloop>=0.0.4; extra == 'runloop'
 Provides-Extra: together

{deepagents_code-0.1.3 → deepagents_code-0.1.4}/deepagents_code/_server_config.py

@@ -152,6 +152,26 @@ class ServerConfig:
     enable_skills: bool = True
     """Enable the skills subsystem (SKILL.md loading and skill tools)."""
 
+    enable_interpreter: bool = False
+    """Enable `CodeInterpreterMiddleware` (`js_eval` REPL) on the main agent.
+
+    Local-mode only; the server graph raises if a sandbox is configured and
+    this flag is `True`.
+    """
+
+    interpreter_ptc: str | list[str] | None = None
+    """Override for `settings.interpreter_ptc`.
+
+    `None` means "fall through to whatever `settings.interpreter_ptc` resolves
+    to from `~/.deepagents/config.toml`". A string is one of `"safe"`/`"all"`;
+    a list is an explicit allowlist of tool names.
+    """
+
+    interpreter_ptc_acknowledge_unsafe: bool = False
+    """Mirror of `settings.interpreter_ptc_acknowledge_unsafe` — required when
+    `interpreter_ptc="all"` is paired with non-`auto_approve` mode.
+    """
+
     sandbox_type: str | None = None
     """Sandbox backend identifier (e.g. `'daytona'`); `None` runs tools on the
     host. `'none'` is normalized to `None` in `__post_init__`."""
@@ -159,6 +179,10 @@ class ServerConfig:
     sandbox_id: str | None = None
     """Existing sandbox ID to attach to; `None` creates a fresh sandbox."""
 
+    sandbox_snapshot_name: str | None = None
+    """Sandbox snapshot name to use or create. LangSmith-only; must be `None`
+    when `sandbox_id` is set."""
+
     sandbox_setup: str | None = None
     """Absolute path to a setup script executed inside the sandbox on first attach."""
 
@@ -226,8 +250,18 @@ class ServerConfig:
             "ENABLE_ASK_USER": str(self.enable_ask_user).lower(),
             "ENABLE_MEMORY": str(self.enable_memory).lower(),
             "ENABLE_SKILLS": str(self.enable_skills).lower(),
+            "ENABLE_INTERPRETER": str(self.enable_interpreter).lower(),
+            "INTERPRETER_PTC": (
+                json.dumps(self.interpreter_ptc)
+                if self.interpreter_ptc is not None
+                else None
+            ),
+            "INTERPRETER_PTC_ACKNOWLEDGE_UNSAFE": str(
+                self.interpreter_ptc_acknowledge_unsafe
+            ).lower(),
             "SANDBOX_TYPE": self.sandbox_type,
             "SANDBOX_ID": self.sandbox_id,
+            "SANDBOX_SNAPSHOT_NAME": self.sandbox_snapshot_name,
             "SANDBOX_SETUP": self.sandbox_setup,
             "CWD": self.cwd,
             "PROJECT_ROOT": self.project_root,
@@ -268,8 +302,14 @@ class ServerConfig:
             enable_ask_user=_read_env_bool("ENABLE_ASK_USER"),
             enable_memory=_read_env_bool("ENABLE_MEMORY", default=True),
             enable_skills=_read_env_bool("ENABLE_SKILLS", default=True),
+            enable_interpreter=_read_env_bool("ENABLE_INTERPRETER"),
+            interpreter_ptc=_read_env_json("INTERPRETER_PTC"),
+            interpreter_ptc_acknowledge_unsafe=_read_env_bool(
+                "INTERPRETER_PTC_ACKNOWLEDGE_UNSAFE"
+            ),
             sandbox_type=_read_env_str("SANDBOX_TYPE"),
             sandbox_id=_read_env_str("SANDBOX_ID"),
+            sandbox_snapshot_name=_read_env_str("SANDBOX_SNAPSHOT_NAME") or None,
             sandbox_setup=_read_env_str("SANDBOX_SETUP"),
             cwd=_read_env_str("CWD"),
             project_root=_read_env_str("PROJECT_ROOT"),
@@ -295,9 +335,13 @@ class ServerConfig:
         shell_allow_list: list[str] | None = None,
         sandbox_type: str = "none",
         sandbox_id: str | None,
+        sandbox_snapshot_name: str | None,
         sandbox_setup: str | None,
         enable_shell: bool,
         enable_ask_user: bool,
+        enable_interpreter: bool = False,
+        interpreter_ptc: str | list[str] | None = None,
+        interpreter_ptc_acknowledge_unsafe: bool = False,
         mcp_config_path: str | None,
         no_mcp: bool,
         trust_project_mcp: bool | None,
@@ -321,9 +365,16 @@ class ServerConfig:
                 server subprocess for `ShellAllowListMiddleware`.
             sandbox_type: Sandbox type.
             sandbox_id: Existing sandbox ID to reuse.
+            sandbox_snapshot_name: Sandbox snapshot name to use or create
+                (langsmith only).
             sandbox_setup: Path to setup script for the sandbox.
             enable_shell: Enable shell execution tools.
             enable_ask_user: Enable ask_user tool.
+            enable_interpreter: Enable `CodeInterpreterMiddleware` on the main
+                agent.
+            interpreter_ptc: Override for `settings.interpreter_ptc`.
+            interpreter_ptc_acknowledge_unsafe: Mirror of
+                `settings.interpreter_ptc_acknowledge_unsafe`.
             mcp_config_path: Path to MCP config.
             no_mcp: Disable MCP.
             trust_project_mcp: Trust project MCP servers.
@@ -344,8 +395,12 @@ class ServerConfig:
             interactive=interactive,
             enable_shell=enable_shell,
             enable_ask_user=enable_ask_user,
+            enable_interpreter=enable_interpreter,
+            interpreter_ptc=interpreter_ptc,
+            interpreter_ptc_acknowledge_unsafe=interpreter_ptc_acknowledge_unsafe,
             sandbox_type=sandbox_type,
             sandbox_id=sandbox_id,
+            sandbox_snapshot_name=sandbox_snapshot_name,
             sandbox_setup=_normalize_path(
                 sandbox_setup, project_context, "sandbox setup"
             ),

deepagents_code-0.1.4/deepagents_code/_startup_error.py

@@ -0,0 +1,45 @@
+"""Stderr marker emission used by the langgraph server graph entry point.
+
+Lives in its own module so unit tests can exercise the marker contract
+without triggering `server_graph.make_graph()` at import time.
+"""
+
+from __future__ import annotations
+
+import logging
+import sys
+import traceback
+
+logger = logging.getLogger(__name__)
+
+STARTUP_ERROR_MARKER = "DEEPAGENTS_STARTUP_ERROR:"
+"""Stderr marker the parent app scans for in `server._extract_startup_error_marker`
+to upgrade an opaque "Server process exited with code N" into a one-line summary.
+Format is `{STARTUP_ERROR_MARKER}{single-line message}`."""
+
+
+def emit_startup_failure(exc: BaseException) -> None:
+    """Report a server graph startup failure to the parent app process.
+
+    Emits two stderr outputs: the full traceback for logs/debugging, then a
+    single-line `{STARTUP_ERROR_MARKER}{type}: {summary}` line that
+    `server._extract_startup_error_marker` parses to upgrade an opaque
+    "Server process exited with code N" into an actionable summary.
+
+    Args:
+        exc: The exception raised during graph initialization.
+    """
+    logger.critical("Failed to initialize server graph", exc_info=exc)
+    print(  # noqa: T201  # stderr fallback — logger may not reach parent process
+        f"Failed to initialize server graph: {exc}\n{traceback.format_exc()}",
+        file=sys.stderr,
+    )
+    # Marker contract is single-line; guard against multi-line/empty `str(exc)`
+    # and include the type so e.g. `ValueError` and `RuntimeError` are
+    # distinguishable in the parent's truncated summary.
+    exc_lines = str(exc).splitlines()
+    summary = exc_lines[0] if exc_lines else "<no message>"
+    print(  # noqa: T201
+        f"{STARTUP_ERROR_MARKER}{type(exc).__name__}: {summary}",
+        file=sys.stderr,
+    )

{deepagents_code-0.1.3 → deepagents_code-0.1.4}/deepagents_code/_version.py

@@ -2,7 +2,7 @@
 
 # Keep the `x-release-please-version` annotation — release-please uses it to
 # bump `__version__` in sync with `pyproject.toml` on every release PR.
-__version__ = "0.1.3"  # x-release-please-version
+__version__ = "0.1.4"  # x-release-please-version
 
 DOCS_URL = "https://docs.langchain.com/oss/python/deepagents/code"
 """URL for `deepagents-code` documentation."""

{deepagents_code-0.1.3 → deepagents_code-0.1.4}/deepagents_code/agent.py

@@ -9,7 +9,7 @@ import shutil
 import tempfile
 import tomllib
 from pathlib import Path
-from typing import TYPE_CHECKING, Any
+from typing import TYPE_CHECKING, Any, cast
 
 from deepagents import create_deep_agent
 from deepagents.backends import CompositeBackend, LocalShellBackend
@@ -198,6 +198,121 @@ class ShellAllowListMiddleware(AgentMiddleware):
         return await handler(request)
 
 
+_INTERPRETER_WRITE_TOOLS: frozenset[str] = frozenset(
+    {"execute", "write_file", "edit_file"}
+)
+"""Tools considered write/shell capable for PTC auditing.
+
+When `interpreter_ptc="all"` resolves to this set, an INFO log names every
+write tool that was included so the audit trail is searchable. The `"safe"`
+preset already excludes them; this is the belt-and-braces check for `"all"`.
+"""
+
+
+def _resolve_ptc_option(
+    ptc: str | bool | list[str],
+    *,
+    tools: Sequence[BaseTool | Callable | dict[str, Any]],
+    acknowledge_unsafe: bool,
+    auto_approve: bool,
+) -> list[str] | None:
+    """Resolve the configured PTC allowlist to a concrete list of tool names.
+
+    Args:
+        ptc: Raw `interpreter_ptc` value from settings or CLI. Accepts
+            `False`/`[]`, `"safe"`, `"all"`, or a list of names.
+        tools: Live tool list given to `create_cli_agent`. Used to validate
+            explicit names, intersect the `"safe"` preset, and enumerate
+            `"all"`.
+        acknowledge_unsafe: Mirrors `settings.interpreter_ptc_acknowledge_unsafe`;
+            required when `ptc="all"` and `auto_approve` is `False`.
+        auto_approve: Whether HITL approval is globally disabled. When `True`,
+            `"all"` does not require `acknowledge_unsafe` because every host
+            tool already runs without prompting.
+
+    Returns:
+        `None` when PTC should be disabled, otherwise a list of tool names
+        suitable for `CodeInterpreterMiddleware(ptc=...)`.
+
+    Raises:
+        ValueError: For unknown names in an explicit list, or for `"all"`
+            without `acknowledge_unsafe` outside of `auto_approve`.
+    """
+    from langchain.tools import BaseTool as _BaseTool
+
+    if ptc is False or ptc is None or ptc == []:
+        return None
+
+    live_names: list[str] = []
+    for tool in tools:
+        if isinstance(tool, _BaseTool):
+            name = tool.name
+            if isinstance(name, str):
+                live_names.append(name)
+        elif isinstance(tool, dict):
+            raw_name = cast("dict[str, Any]", tool).get("name")
+            if isinstance(raw_name, str):
+                live_names.append(raw_name)
+        else:
+            attr = getattr(tool, "name", None)
+            if isinstance(attr, str):
+                live_names.append(attr)
+    live_set: set[str] = set(live_names)
+
+    if isinstance(ptc, str):
+        normalized = ptc.strip().lower()
+        if normalized == "safe":
+            from deepagents_code.config import INTERPRETER_PTC_SAFE_PRESET
+
+            selected = sorted(INTERPRETER_PTC_SAFE_PRESET & live_set)
+            dropped = sorted(INTERPRETER_PTC_SAFE_PRESET - live_set)
+            if dropped:
+                logger.debug(
+                    "interpreter_ptc='safe' preset members not present in toolset: %s",
+                    dropped,
+                )
+            return selected
+        if normalized == "all":
+            if not auto_approve and not acknowledge_unsafe:
+                msg = (
+                    "interpreter_ptc='all' exposes every host tool to PTC "
+                    "calls that bypass HITL approval. Set "
+                    "interpreter_ptc_acknowledge_unsafe=True (or use "
+                    "auto_approve=True) to opt in."
+                )
+                raise ValueError(msg)
+            included = sorted(live_set)
+            write_included = sorted(_INTERPRETER_WRITE_TOOLS & live_set)
+            if write_included:
+                logger.info(
+                    "interpreter_ptc='all' includes write/shell tools: %s",
+                    write_included,
+                )
+            return included
+        msg = (
+            f"Invalid interpreter_ptc string {ptc!r}; expected 'safe', 'all', "
+            "or a list of tool names."
+        )
+        raise ValueError(msg)
+
+    if isinstance(ptc, list):
+        unknown = [name for name in ptc if name not in live_set]
+        if unknown:
+            available = ", ".join(sorted(live_set)) or "<none>"
+            msg = (
+                "Unknown tool names in interpreter_ptc: "
+                f"{sorted(set(unknown))}. Available tools: {available}."
+            )
+            raise ValueError(msg)
+        return list(ptc)
+
+    msg = (
+        "interpreter_ptc must be False, 'safe', 'all', or a list of tool names; "
+        f"got {type(ptc).__name__}."
+    )
+    raise ValueError(msg)
+
+
 def load_async_subagents(config_path: Path | None = None) -> list[AsyncSubAgent]:
     """Load async subagent definitions from `config.toml`.
 
@@ -924,6 +1039,7 @@ def create_cli_agent(
     enable_memory: bool = True,
     enable_skills: bool = True,
     enable_shell: bool = True,
+    enable_interpreter: bool = False,
     checkpointer: BaseCheckpointSaver | None = None,
     mcp_server_info: list[MCPServerInfo] | None = None,
     cwd: str | Path | None = None,
@@ -980,6 +1096,27 @@ def create_cli_agent(
         enable_skills: Enable `SkillsMiddleware` for custom agent skills
         enable_shell: Enable shell execution via `LocalShellBackend`
             (only in local mode). When enabled, the `execute` tool is available.
+        enable_interpreter: Wire `CodeInterpreterMiddleware` from
+            `langchain-quickjs` into the main agent.
+
+            Local-mode only — passing a non-`None` `sandbox` while
+            `enable_interpreter=True` raises `ValueError`. Subagents do not
+            receive the interpreter in v1.
+
+            PTC (`tools.*` host bridge) calls bypass `interrupt_on`/HITL
+            approval, so `settings.interpreter_ptc` is the only effective
+            control over which host tools can be invoked from inside the
+            REPL. `js_eval` itself is intentionally not gated by HITL —
+            per-call approval would be unusably noisy and would not block
+            PTC fan-out anyway. The `"safe"` preset is therefore restricted
+            to tools that are already non-HITL outside the REPL (read-only
+            file inspection); exposing HITL-gated tools — network fetch,
+            subagent dispatch, shell, file writes — requires an explicit
+            list or `interpreter_ptc="all"` with
+            `interpreter_ptc_acknowledge_unsafe=True`.
+
+            Requires the `quickjs` optional extra
+            (`langchain-quickjs>=0.1.2,<0.2.0`).
         checkpointer: Optional checkpointer for session persistence.
             When `None`, the graph is compiled without a checkpointer.
         mcp_server_info: MCP server metadata to surface in the system prompt.
@@ -998,6 +1135,12 @@ def create_cli_agent(
             - `agent_graph`: Configured LangGraph Pregel instance ready
                 for execution
             - `composite_backend`: `CompositeBackend` for file operations
+
+    Raises:
+        ValueError: When `enable_interpreter=True` is paired with a
+            non-`None` `sandbox`, when `settings.interpreter_ptc` contains
+            unknown tool names, or when `interpreter_ptc="all"` is used
+            without `auto_approve` or `interpreter_ptc_acknowledge_unsafe`.
     """
     tools = tools or []
     effective_cwd = (
@@ -1201,6 +1344,37 @@ def create_cli_agent(
         # Note: Shell middleware not used in sandbox mode
         # File operations and execute tool are provided by the sandbox backend
 
+    if enable_interpreter:
+        if sandbox is not None:
+            msg = (
+                "enable_interpreter=True is not supported with a remote "
+                "sandbox in this release. Disable the sandbox or unset "
+                "enable_interpreter."
+            )
+            raise ValueError(msg)
+        # Lazy import keeps `dcode -v` fast — see AGENTS.md startup-perf rule.
+        from langchain_quickjs import CodeInterpreterMiddleware, PTCOption
+
+        ptc_names = _resolve_ptc_option(
+            settings.interpreter_ptc,
+            tools=tools,
+            acknowledge_unsafe=settings.interpreter_ptc_acknowledge_unsafe,
+            auto_approve=auto_approve,
+        )
+        ptc_option: PTCOption | None = (
+            cast("PTCOption", list(ptc_names)) if ptc_names is not None else None
+        )
+        agent_middleware.append(
+            CodeInterpreterMiddleware(
+                tool_name="js_eval",
+                timeout=settings.interpreter_timeout_seconds,
+                memory_limit=settings.interpreter_memory_limit_mb * 1024 * 1024,
+                max_ptc_calls=settings.interpreter_max_ptc_calls,
+                max_result_chars=settings.interpreter_max_result_chars,
+                ptc=ptc_option,
+            )
+        )
+
     # Local context middleware (git info, directory tree, etc.).
     if isinstance(backend, (_ExecutableBackend, _AsyncExecutableBackend)):
         agent_middleware.append(

{deepagents_code-0.1.3 → deepagents_code-0.1.4}/deepagents_code/app.py

@@ -3446,7 +3446,13 @@ class DeepAgentsApp(App):
         Custom themes use their stored registry colors; built-in Textual themes
         resolve colors from the active app theme. Terminal write failures are
         logged and swallowed because the OSC background sync is cosmetic.
+
+        ANSI themes intentionally skip this step so the terminal's native
+        background is preserved.
         """
+        if self.theme in {"ansi-dark", "ansi-light"}:
+            return
+
         from deepagents_code.terminal_escape import set_terminal_background
 
         entry = theme.get_registry().get(self.theme)
@@ -4918,7 +4924,14 @@ class DeepAgentsApp(App):
         Args:
             command: The raw command text (displayed as user message).
         """
-        from deepagents_code.config import build_langsmith_thread_url
+        from deepagents_code.config import (
+            LangSmithApiError,
+            LangSmithImportError,
+            LangSmithLookupTimeoutError,
+            _assemble_langsmith_thread_url,
+            fetch_langsmith_project_url_or_raise,
+            get_langsmith_project_name,
+        )
 
         if not self._session_state:
             await self._mount_message(UserMessage(command))
@@ -4926,15 +4939,18 @@ class DeepAgentsApp(App):
             return
         thread_id = self._session_state.thread_id
         try:
-            url = await asyncio.to_thread(build_langsmith_thread_url, thread_id)
+            project_name = await asyncio.to_thread(get_langsmith_project_name)
         except Exception:
-            logger.exception("Failed to build LangSmith thread URL for %s", thread_id)
+            logger.exception(
+                "Failed to resolve LangSmith project name for thread %s",
+                thread_id,
+            )
             await self._mount_message(UserMessage(command))
             await self._mount_message(
-                AppMessage("Failed to resolve LangSmith thread URL."),
+                AppMessage("Failed to resolve LangSmith project name."),
             )
             return
-        if not url:
+        if not project_name:
             await self._mount_message(UserMessage(command))
             await self._mount_message(
                 AppMessage(
@@ -4943,6 +4959,61 @@ class DeepAgentsApp(App):
                 ),
             )
             return
+        try:
+            project_url = await asyncio.to_thread(
+                fetch_langsmith_project_url_or_raise, project_name
+            )
+        except LangSmithImportError:
+            logger.warning(
+                "langsmith package not installed; cannot resolve thread URL for %s",
+                thread_id,
+            )
+            await self._mount_message(UserMessage(command))
+            await self._mount_message(
+                AppMessage(
+                    "The `langsmith` package is not installed. "
+                    "Install it with `pip install langsmith` to enable `/trace`.",
+                ),
+            )
+            return
+        except LangSmithLookupTimeoutError:
+            logger.warning(
+                "LangSmith project URL lookup timed out for thread %s",
+                thread_id,
+            )
+            await self._mount_message(UserMessage(command))
+            await self._mount_message(
+                AppMessage(
+                    "Could not reach LangSmith to resolve the thread URL. "
+                    "Check your network connection and try again.",
+                ),
+            )
+            return
+        except LangSmithApiError as exc:
+            logger.warning(
+                "LangSmith API call failed while resolving thread URL for %s: %s",
+                thread_id,
+                exc,
+            )
+            await self._mount_message(UserMessage(command))
+            await self._mount_message(
+                AppMessage(
+                    f"LangSmith rejected the project lookup: {exc}. "
+                    "Verify LANGSMITH_API_KEY and the project name are correct.",
+                ),
+            )
+            return
+        except Exception:
+            logger.exception(
+                "Failed to fetch LangSmith project URL for thread %s",
+                thread_id,
+            )
+            await self._mount_message(UserMessage(command))
+            await self._mount_message(
+                AppMessage("Failed to resolve LangSmith thread URL."),
+            )
+            return
+        url = _assemble_langsmith_thread_url(project_url, thread_id)
 
         def _open_browser() -> None:
             try: