codex-autorunner 0.1.2__tar.gz → 1.1.0__tar.gz

codex_autorunner-1.1.0/PKG-INFO

@@ -0,0 +1,154 @@
+Metadata-Version: 2.4
+Name: codex-autorunner
+Version: 1.1.0
+Summary: Codex autorunner CLI per DESIGN-V1
+Author: Codex
+License: MIT License
+        
+        Copyright (c) 2025 David Zhang
+        
+        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/Git-on-my-level/codex-autorunner
+Project-URL: Repository, https://github.com/Git-on-my-level/codex-autorunner
+Project-URL: Issues, https://github.com/Git-on-my-level/codex-autorunner/issues
+Keywords: codex,automation,agent,cli,fastapi
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3 :: Only
+Classifier: Operating System :: OS Independent
+Requires-Python: >=3.9
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: typer>=0.9
+Requires-Dist: click<8.2
+Requires-Dist: pyyaml>=6.0
+Requires-Dist: fastapi>=0.111
+Requires-Dist: uvicorn[standard]>=0.30
+Requires-Dist: ptyprocess>=0.7
+Requires-Dist: python-multipart>=0.0.9
+Requires-Dist: python-dotenv>=1.0
+Requires-Dist: httpx>=0.27
+Requires-Dist: tenacity>=8.0
+Provides-Extra: dev
+Requires-Dist: black==25.11.0; extra == "dev"
+Requires-Dist: mypy>=1.10; extra == "dev"
+Requires-Dist: pytest>=7.0; extra == "dev"
+Requires-Dist: pytest-asyncio>=0.23; extra == "dev"
+Requires-Dist: pytest-timeout>=2.0; extra == "dev"
+Requires-Dist: ruff>=0.5.0; extra == "dev"
+Requires-Dist: types-PyYAML; extra == "dev"
+Provides-Extra: telegram
+Requires-Dist: httpx>=0.27; extra == "telegram"
+Provides-Extra: voice
+Requires-Dist: httpx>=0.27; extra == "voice"
+Requires-Dist: python-multipart>=0.0.9; extra == "voice"
+Provides-Extra: github
+Dynamic: license-file
+
+# CAR (codex-autorunner)
+[![PyPI](https://img.shields.io/pypi/v/codex-autorunner.svg)](https://pypi.org/project/codex-autorunner/)
+
+CAR provides a set of low-opinion agent coordination tools for you to run long complex implementations using the agents you already love.
+
+What this looks like in practice:
+- You write a plan, or generate a plan by chatting with your favorite AI
+- You convert the plan (or ask an AI to convert it for you) into CAR compatible tickets (markdown with some frontmatter)
+- Go off and do something else, no need to babysit the agents, they will notify you if they need your input
+
+![CAR Web Hub Screenshot](docs/screenshots/hub.png)
+
+## How it works
+CAR is very simple. At it's core, CAR is a state machine which checks to see if there are any incomplete tickets. If yes, pick the next one and run it against an agent. Tickets can be pre-populated by the user, but agents can also write tickets. _Tickets are the control plane for CAR_.
+
+When each agent wakes up, it gets knowledge about CAR and how to operate within CAR, a pre-defined set of context (workspace files), the current ticket, and optionally the final output of the previous agent. This simple loop ensures that agents know enough to use CAR while also focusing them on the task at hand.
+
+## Philosophy
+The philosophy behind CAR is to let the agents do what they do best, and get out of their way. CAR is _very bitter-lesson-pilled_. As models and agents get more powerful, CAR should serve as a form of leverage, and avoid constraining models and their harnesses. This is why we treat the filesystem as the first class data plane and utilize tools and languages the models are already very familiar with (git, python).
+
+CAR treats tickets as the control plane and models as the execution layer. This means that we rely on agents to follow the instructions written in the tickets. If you use a sufficiently weak model, CAR may not work well for you. CAR is an amplifier for agent capabilities. Agents who like to scope creep (create too many new tickets) or reward hack (mark a ticket as done despite it being incomplete) are not a good fit for CAR.
+
+## Interaction patterns
+CAR's core is a set of python functions surfaced as a CLI, operating on a file system. There are current 2 UIs built on top of this core.
+
+### Web UI
+The web UI is the main control plane for CAR. From here you can set up new repositories or clone existing ones, chat with agents using their TUI, and run the ticket autorunner. There are many quality-of-life features like Whisper integration, editing documents by chatting with AI (useful for mobile), viewing usage analytics, and much more. The Web UI is the most full featured user-facing interface and a good starting point for trying out CAR.
+
+I recommend serving the web UI over Tailscale. There is an auth token option but the system is not very battle tested.
+
+### Telegram
+Telegram is the "on-the-go" and notification hub for CAR. From here you can kick off and monitor existing tickets, set up new tickets, and chat with agents. Your primary UX here is asking the agent to do things for you rather than you doing it yourself like you would on the web UI. This is great for on-the-go work, but it doesn't have full feature parity with the web UI.
+
+## Quickstart
+
+The fastest way to get started is to pass [this setup guide](docs/AGENT_SETUP_GUIDE.md) to your favorite AI agent. The agent will walk you through installation and configuration interactively based on your environment.
+
+**TL;DR for the impatient:**
+
+# Install
+```
+pipx install codex-autorunner
+```
+# Initialize in your repo
+```
+cd /path/to/your/repo
+car init
+```
+# Verify setup
+```
+car doctor
+```
+# Create a ticket and run
+```
+car run
+```
+
+## Supported models
+CAR currently supports:
+- Codex
+- Opencode
+
+CAR is built to easily integrate any reasonable agent built for Agent Client Protocol (ACP). If you would like to see your agent supported, please reach out or open a PR.
+
+## Examples
+Build out complex features and products by providing a series of tickets assigned to various agents.
+![CAR Tickets in Progress Screenshot](docs/screenshots/tickets-in-progress.png)
+
+Tickets are just markdown files that both you and the agent can edit.
+![CAR Ticket Markdown Screenshot](docs/screenshots/ticket-markdown.png)
+
+You don't have to babysit the agents, they inbox you or ping you on Telegram.
+![CAR Inbox Screenshot](docs/screenshots/inbox.png)
+
+You can collaborate with the agents in a shared workspace, independent of the codebase. Drop context there, extract artifacts, it's like a shared scratchpad.
+![CAR Workspace Screenshot](docs/screenshots/workspace.png)
+
+All core workspace documents are also just markdown files, so you and the agent can easily edit them.
+![CAR Workspace New MD Screenshot](docs/screenshots/workspace-new-md.png)
+
+If you need to do something more custom or granular, you can use your favorite agent TUIs in the built-in terminal.
+![CAR Terminal Codex Screenshot](docs/screenshots/terminal-codex.png)
+![CAR Terminal Opencode Screenshot](docs/screenshots/terminal-opencode.png)
+
+On the go? The web UI is mobile responsive, or if you prefer you can type or voice chat with your agents on Telegram.
+![CAR Telegram Media Voice Screenshot](docs/screenshots/telegram-media-voice.PNG)
+![CAR Telegram Media Image Screenshot](docs/screenshots/telegram-media-image.PNG)
+
+## Star history
+[![Star History Chart](https://api.star-history.com/svg?repos=Git-on-my-level/codex-autorunner&type=Date)](https://star-history.com/#Git-on-my-level/codex-autorunner&Date)

codex_autorunner-1.1.0/README.md

@@ -0,0 +1,90 @@
+# CAR (codex-autorunner)
+[![PyPI](https://img.shields.io/pypi/v/codex-autorunner.svg)](https://pypi.org/project/codex-autorunner/)
+
+CAR provides a set of low-opinion agent coordination tools for you to run long complex implementations using the agents you already love.
+
+What this looks like in practice:
+- You write a plan, or generate a plan by chatting with your favorite AI
+- You convert the plan (or ask an AI to convert it for you) into CAR compatible tickets (markdown with some frontmatter)
+- Go off and do something else, no need to babysit the agents, they will notify you if they need your input
+
+![CAR Web Hub Screenshot](docs/screenshots/hub.png)
+
+## How it works
+CAR is very simple. At it's core, CAR is a state machine which checks to see if there are any incomplete tickets. If yes, pick the next one and run it against an agent. Tickets can be pre-populated by the user, but agents can also write tickets. _Tickets are the control plane for CAR_.
+
+When each agent wakes up, it gets knowledge about CAR and how to operate within CAR, a pre-defined set of context (workspace files), the current ticket, and optionally the final output of the previous agent. This simple loop ensures that agents know enough to use CAR while also focusing them on the task at hand.
+
+## Philosophy
+The philosophy behind CAR is to let the agents do what they do best, and get out of their way. CAR is _very bitter-lesson-pilled_. As models and agents get more powerful, CAR should serve as a form of leverage, and avoid constraining models and their harnesses. This is why we treat the filesystem as the first class data plane and utilize tools and languages the models are already very familiar with (git, python).
+
+CAR treats tickets as the control plane and models as the execution layer. This means that we rely on agents to follow the instructions written in the tickets. If you use a sufficiently weak model, CAR may not work well for you. CAR is an amplifier for agent capabilities. Agents who like to scope creep (create too many new tickets) or reward hack (mark a ticket as done despite it being incomplete) are not a good fit for CAR.
+
+## Interaction patterns
+CAR's core is a set of python functions surfaced as a CLI, operating on a file system. There are current 2 UIs built on top of this core.
+
+### Web UI
+The web UI is the main control plane for CAR. From here you can set up new repositories or clone existing ones, chat with agents using their TUI, and run the ticket autorunner. There are many quality-of-life features like Whisper integration, editing documents by chatting with AI (useful for mobile), viewing usage analytics, and much more. The Web UI is the most full featured user-facing interface and a good starting point for trying out CAR.
+
+I recommend serving the web UI over Tailscale. There is an auth token option but the system is not very battle tested.
+
+### Telegram
+Telegram is the "on-the-go" and notification hub for CAR. From here you can kick off and monitor existing tickets, set up new tickets, and chat with agents. Your primary UX here is asking the agent to do things for you rather than you doing it yourself like you would on the web UI. This is great for on-the-go work, but it doesn't have full feature parity with the web UI.
+
+## Quickstart
+
+The fastest way to get started is to pass [this setup guide](docs/AGENT_SETUP_GUIDE.md) to your favorite AI agent. The agent will walk you through installation and configuration interactively based on your environment.
+
+**TL;DR for the impatient:**
+
+# Install
+```
+pipx install codex-autorunner
+```
+# Initialize in your repo
+```
+cd /path/to/your/repo
+car init
+```
+# Verify setup
+```
+car doctor
+```
+# Create a ticket and run
+```
+car run
+```
+
+## Supported models
+CAR currently supports:
+- Codex
+- Opencode
+
+CAR is built to easily integrate any reasonable agent built for Agent Client Protocol (ACP). If you would like to see your agent supported, please reach out or open a PR.
+
+## Examples
+Build out complex features and products by providing a series of tickets assigned to various agents.
+![CAR Tickets in Progress Screenshot](docs/screenshots/tickets-in-progress.png)
+
+Tickets are just markdown files that both you and the agent can edit.
+![CAR Ticket Markdown Screenshot](docs/screenshots/ticket-markdown.png)
+
+You don't have to babysit the agents, they inbox you or ping you on Telegram.
+![CAR Inbox Screenshot](docs/screenshots/inbox.png)
+
+You can collaborate with the agents in a shared workspace, independent of the codebase. Drop context there, extract artifacts, it's like a shared scratchpad.
+![CAR Workspace Screenshot](docs/screenshots/workspace.png)
+
+All core workspace documents are also just markdown files, so you and the agent can easily edit them.
+![CAR Workspace New MD Screenshot](docs/screenshots/workspace-new-md.png)
+
+If you need to do something more custom or granular, you can use your favorite agent TUIs in the built-in terminal.
+![CAR Terminal Codex Screenshot](docs/screenshots/terminal-codex.png)
+![CAR Terminal Opencode Screenshot](docs/screenshots/terminal-opencode.png)
+
+On the go? The web UI is mobile responsive, or if you prefer you can type or voice chat with your agents on Telegram.
+![CAR Telegram Media Voice Screenshot](docs/screenshots/telegram-media-voice.PNG)
+![CAR Telegram Media Image Screenshot](docs/screenshots/telegram-media-image.PNG)
+
+## Star history
+[![Star History Chart](https://api.star-history.com/svg?repos=Git-on-my-level/codex-autorunner&type=Date)](https://star-history.com/#Git-on-my-level/codex-autorunner&Date)

{codex_autorunner-0.1.2 → codex_autorunner-1.1.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "codex-autorunner"
-version = "0.1.2"
+version = "1.1.0"
 description = "Codex autorunner CLI per DESIGN-V1"
 readme = "README.md"
 license = {file = "LICENSE"}
@@ -82,14 +82,21 @@ completions = true
 cache_dir = ".codex-autorunner/pytest-cache"
 asyncio_mode = "auto"
 asyncio_default_fixture_loop_scope = "function"
+norecursedirs = ["worktrees", ".git", ".venv", ".codex-autorunner"]
 markers = [
     "integration: marks tests as integration tests that require external dependencies (deselect with '-m \"not integration\"')",
     "timeout(seconds): per-test timeout (provided by pytest-timeout)",
 ]
+filterwarnings = [
+    # Cancelled auto-restart tasks can emit noisy unraisable warnings when the event
+    # loop shuts down; we explicitly clean these up, so silence the residual notice.
+    "ignore::pytest.PytestUnraisableExceptionWarning",
+]
 
 [tool.ruff]
 line-length = 88
 target-version = "py39"
+src = ["src"]
 
 [tool.ruff.lint]
 select = ["E", "F", "W", "I", "B"]
@@ -100,8 +107,11 @@ ignore = ["E501"]
 "src/codex_autorunner/pty_session.py" = ["F403"]
 "src/codex_autorunner/static_assets.py" = ["F403"]
 "src/codex_autorunner/cli.py" = ["B008"]
+"src/codex_autorunner/surfaces/cli/cli.py" = ["B008"]
 "src/codex_autorunner/routes/terminal_images.py" = ["B008"]
+"src/codex_autorunner/surfaces/web/routes/terminal_images.py" = ["B008"]
 "src/codex_autorunner/routes/voice.py" = ["B008"]
+"src/codex_autorunner/surfaces/web/routes/voice.py" = ["B008"]
 
 [tool.mypy]
 python_version = "3.9"

codex_autorunner-1.1.0/src/codex_autorunner/__init__.py

@@ -0,0 +1,14 @@
+"""Codex autorunner package."""
+
+__all__ = [
+    "cli",
+    "core",
+    "integrations",
+    "routes",
+    "server",
+    "surfaces",
+    "surfaces.web.routes",
+    "surfaces.web",
+    "voice",
+    "web",
+]

codex_autorunner-1.1.0/src/codex_autorunner/__main__.py

@@ -0,0 +1,4 @@
+from .cli import app
+
+if __name__ == "__main__":
+    app()

{codex_autorunner-0.1.2 → codex_autorunner-1.1.0}/src/codex_autorunner/agents/codex/harness.py

@@ -3,7 +3,7 @@ from __future__ import annotations
 from pathlib import Path
 from typing import Any, AsyncIterator, Optional
 
-from ...core.app_server_events import AppServerEventBuffer
+from ...integrations.app_server.event_buffer import AppServerEventBuffer
 from ...integrations.app_server.supervisor import WorkspaceAppServerSupervisor
 from ..base import AgentHarness
 from ..types import AgentId, ConversationRef, ModelCatalog, ModelSpec, TurnRef

{codex_autorunner-0.1.2 → codex_autorunner-1.1.0}/src/codex_autorunner/agents/opencode/client.py

@@ -4,6 +4,7 @@ import asyncio
 import dataclasses
 import json
 import logging
+import re
 from typing import Any, AsyncIterator, Iterable, Optional
 
 import httpx
@@ -48,7 +49,18 @@ def _normalize_sse_event(event: SSEEvent) -> SSEEvent:
         payload_obj = None
 
     if isinstance(payload_obj, dict) and isinstance(payload_obj.get("payload"), dict):
-        payload_obj = payload_obj["payload"]
+        outer = payload_obj
+        inner = dict(outer.get("payload") or {})
+        if "type" not in inner and isinstance(outer.get("type"), str):
+            inner["type"] = outer["type"]
+        for key in ("sessionID", "sessionId", "session_id"):
+            if key in outer and key not in inner:
+                inner[key] = outer[key]
+        if "session" in outer and "session" not in inner:
+            inner["session"] = outer["session"]
+        if "properties" in outer and "properties" not in inner:
+            inner["properties"] = outer["properties"]
+        payload_obj = inner
 
     if isinstance(payload_obj, dict):
         payload_type = payload_obj.get("type")
@@ -505,49 +517,70 @@ class OpenCodeClient:
 
     async def fetch_openapi_spec(self) -> dict[str, Any]:
         """Fetch OpenAPI spec from /doc endpoint for capability negotiation."""
-        async with self._client.stream("GET", "/doc") as response:
-            response.raise_for_status()
-            content = response.content
-            try:
-                spec = json.loads(content) if content else {}
-                log_event(
-                    self._logger,
-                    logging.INFO,
-                    "opencode.openapi.fetched",
-                    paths=len(spec.get("paths", {})) if isinstance(spec, dict) else 0,
-                    has_components=(
-                        "components" in spec if isinstance(spec, dict) else False
-                    ),
-                )
-                return spec
-            except Exception as exc:
-                log_event(
-                    self._logger,
-                    logging.WARNING,
-                    "opencode.openapi.parse_failed",
-                    exc=exc,
-                )
-                raise OpenCodeProtocolError(
-                    f"Failed to parse OpenAPI spec: {exc}",
-                    status_code=response.status_code,
-                    content_type=(
-                        response.headers.get("content-type") if response else None
-                    ),
-                ) from exc
+        response = await self._client.get("/doc")
+        response.raise_for_status()
+        content = response.content
+        try:
+            spec = json.loads(content) if content else {}
+            log_event(
+                self._logger,
+                logging.INFO,
+                "opencode.openapi.fetched",
+                paths=len(spec.get("paths", {})) if isinstance(spec, dict) else 0,
+                has_components=(
+                    "components" in spec if isinstance(spec, dict) else False
+                ),
+            )
+            return spec
+        except Exception as exc:
+            log_event(
+                self._logger,
+                logging.WARNING,
+                "opencode.openapi.parse_failed",
+                exc=exc,
+            )
+            raise OpenCodeProtocolError(
+                f"Failed to parse OpenAPI spec: {exc}",
+                status_code=response.status_code,
+                content_type=(
+                    response.headers.get("content-type") if response else None
+                ),
+            ) from exc
 
     def has_endpoint(
         self, openapi_spec: dict[str, Any], method: str, path: str
     ) -> bool:
-        """Check if endpoint is available in OpenAPI spec."""
+        """Check if endpoint is available in OpenAPI spec.
+
+        The OpenAPI spec sometimes uses different template parameter names (e.g.,
+        `{sessionID}` vs `{session_id}`). We normalize templates before matching so
+        capability detection does not depend on placeholder spelling.
+        """
         if not isinstance(openapi_spec, dict):
             return False
         paths = openapi_spec.get("paths", {})
         if not isinstance(paths, dict):
             return False
-        path_info = paths.get(path)
-        if not isinstance(path_info, dict):
-            return False
-        return method in path_info
+
+        target = _normalize_template_path(path)
+        method = method.lower()
+
+        for candidate_path, info in paths.items():
+            if not isinstance(info, dict):
+                continue
+            if _normalize_template_path(candidate_path) != target:
+                continue
+            if method in info:
+                return True
+        return False
+
+
+def _normalize_template_path(path: str) -> str:
+    """Collapse template placeholders to a canonical form.
+
+    Example: `/session/{sessionID}/prompt_async` -> `/session/{}/prompt_async`
+    """
+    return re.sub(r"{[^/]+}", "{}", path)
 
 
 __all__ = ["OpenCodeClient", "OpenCodeProtocolError", "OpenCodeApiProfile"]

codex_autorunner-1.1.0/src/codex_autorunner/agents/opencode/constants.py

@@ -0,0 +1,3 @@
+DEFAULT_TICKET_MODEL = "zai-coding-plan/glm-4.7"
+
+__all__ = ["DEFAULT_TICKET_MODEL"]

{codex_autorunner-0.1.2 → codex_autorunner-1.1.0}/src/codex_autorunner/agents/opencode/harness.py

@@ -6,9 +6,10 @@ import logging
 from pathlib import Path
 from typing import Any, AsyncIterator, Optional
 
-from ...core.app_server_events import format_sse
+from ...integrations.app_server.event_buffer import format_sse
 from ..base import AgentHarness
 from ..types import AgentId, ConversationRef, ModelCatalog, ModelSpec, TurnRef
+from .constants import DEFAULT_TICKET_MODEL
 from .runtime import (
     build_turn_id,
     extract_session_id,
@@ -168,6 +169,8 @@ class OpenCodeHarness(AgentHarness):
         sandbox_policy: Optional[Any],
     ) -> TurnRef:
         client = await self._supervisor.get_client(workspace_root)
+        if model is None:
+            model = DEFAULT_TICKET_MODEL
         model_payload = split_model_id(model)
         await client.prompt_async(
             conversation_id,
@@ -192,6 +195,8 @@ class OpenCodeHarness(AgentHarness):
         sandbox_policy: Optional[Any],
     ) -> TurnRef:
         client = await self._supervisor.get_client(workspace_root)
+        if model is None:
+            model = DEFAULT_TICKET_MODEL
         arguments = prompt if prompt else ""
 
         async def _send_review() -> None:

{codex_autorunner-0.1.2 → codex_autorunner-1.1.0}/src/codex_autorunner/agents/opencode/logging.py

@@ -66,11 +66,6 @@ class OpenCodeEventFormatter:
             for line in complete_lines:
                 if line.strip():
                     lines.append(f"**{line.strip()}**")
-
-            remaining = coalescer.get_buffer()
-            if remaining and remaining.strip():
-                lines.append(f"**{remaining.strip()}**")
-                coalescer.clear()
         return lines
 
     def _format_tool_part(self, part: dict[str, Any]) -> list[str]:
@@ -120,6 +115,27 @@ class OpenCodeEventFormatter:
             lines.append("exec")
             lines.append(f"tool: {tool_name}")
 
+        input_preview: Optional[str] = None
+        for key in ("input", "command", "cmd", "script"):
+            value = part.get(key)
+            if isinstance(value, str) and value.strip():
+                input_preview = value.strip()
+                break
+        if input_preview is None:
+            args = part.get("args") or part.get("arguments") or part.get("params")
+            if isinstance(args, dict):
+                for key in ("command", "cmd", "script", "input"):
+                    value = args.get(key)
+                    if isinstance(value, str) and value.strip():
+                        input_preview = value.strip()
+                        break
+            elif isinstance(args, str) and args.strip():
+                input_preview = args.strip()
+        if input_preview:
+            if len(input_preview) > 240:
+                input_preview = input_preview[:240] + "…"
+            lines.append(f"input: {input_preview}")
+
         return lines
 
     def _format_patch_part(self, part: dict[str, Any]) -> list[str]:

{codex_autorunner-0.1.2 → codex_autorunner-1.1.0}/src/codex_autorunner/agents/opencode/run_prompt.py

@@ -112,6 +112,7 @@ async def run_opencode_prompt(
             client,
             session_id=session_id,
             workspace_path=config.workspace_root,
+            model_payload=model_payload,
             permission_policy=permission_policy,
             should_stop=should_stop,
             ready_event=ready_event,