python-codex 0.0.1__py3-none-any.whl → 0.1.1__py3-none-any.whl

pycodex/cli.py

@@ -0,0 +1,705 @@
+from __future__ import annotations
+
+import atexit
+import argparse
+import asyncio
+import json
+import os
+import shlex
+import sys
+import tempfile
+from dataclasses import asdict, replace
+from pathlib import Path
+from typing import Literal, Sequence
+
+from .agent import AgentLoop
+from .collaboration import DEFAULT_COLLABORATION_MODE, CollaborationMode
+from .context import ContextManager
+from .model import ResponsesModelClient, ResponsesProviderConfig
+from .portable import bootstrap_called_home, upload_codex_home
+from .protocol import AgentEvent
+from .runtime import AgentRuntime
+from .runtime_services import RuntimeEnvironment, create_runtime_environment
+from .utils import CliSessionView, load_codex_dotenv
+from responses_server import launch_chat_completion_compat_server
+
+EXIT_COMMANDS = {"/exit", "/quit"}
+HISTORY_COMMAND = "/history"
+TITLE_COMMAND = "/title"
+MODEL_COMMAND = "/model"
+QUEUE_COMMAND = "/queue"
+CliSessionMode = Literal["exec", "tui"]
+LOCAL_RESPONSES_SERVER_API_KEY_ENV = "PYCODEX_LOCAL_RESPONSES_SERVER_KEY"
+CLI_ORIGINATOR = "codex-tui"
+
+
+def configure_loguru() -> None:
+    try:
+        from loguru import logger
+    except ImportError:  # pragma: no cover - dependency may be absent in minimal envs
+        return
+
+    logger.remove()
+    log_path = os.environ.get("PYCODEX_DEBUG_LOG", "").strip()
+    if log_path:
+        logger.add(log_path, level="DEBUG")
+        return
+
+    if os.environ.get("PYCODEX_DEBUG_STDERR", "").strip().lower() in {
+        "1",
+        "true",
+        "yes",
+        "on",
+    }:
+        logger.add(sys.stderr, level="DEBUG")
+
+
+def build_parser() -> argparse.ArgumentParser:
+    parser = argparse.ArgumentParser(
+        prog="pycodex",
+        description="Minimal Codex-style local CLI backed by ~/.codex/config.toml.",
+    )
+    parser.add_argument(
+        "prompt", nargs="*", help="Prompt text. If omitted, read from stdin."
+    )
+    parser.add_argument(
+        "--put",
+        default=None,
+        metavar="PATH@SERVER",
+        help=(
+            "Upload a Codex home using `--put @host:port` or "
+            "`--put /path/.codex@host:port`."
+        ),
+    )
+    parser.add_argument(
+        "--call",
+        default=None,
+        help=(
+            "Download and use a stored Codex home via <secret>-<call_id>@<host:port>."
+        ),
+    )
+    parser.add_argument(
+        "--config",
+        default=str(Path.home() / ".codex" / "config.toml"),
+        help="Path to Codex config.toml.",
+    )
+    parser.add_argument(
+        "--profile",
+        default=None,
+        help="Optional profile name from config.toml.",
+    )
+    parser.add_argument(
+        "--vllm-endpoint",
+        default=None,
+        help=(
+            "Optional base URL for a chat-completions-backed vLLM server. "
+            "When set, pycodex starts a local responses compat server for this "
+            "session and appends /v1 if the path is empty."
+        ),
+    )
+    parser.add_argument(
+        "--use-chat-completion",
+        default=False,
+        action="store_true",
+        help=(
+            "When set, pycodex starts a local responses compat server for this session."
+        ),
+    )
+    parser.add_argument(
+        "--system-prompt",
+        default=None,
+        help="Optional base instructions override passed to the model.",
+    )
+    parser.add_argument(
+        "--timeout-seconds",
+        type=float,
+        default=120.0,
+        help="HTTP timeout for one model call.",
+    )
+    parser.add_argument(
+        "--json",
+        action="store_true",
+        help="Print the full TurnResult as JSON.",
+    )
+    return parser
+
+
+def should_run_interactive(prompt_parts: Sequence[str], stdin_is_tty: bool) -> bool:
+    return not prompt_parts and stdin_is_tty
+
+
+def resolve_prompt_text(prompt_parts: Sequence[str]) -> str:
+    if prompt_parts:
+        return " ".join(prompt_parts).strip()
+
+    if not sys.stdin.isatty():
+        prompt_text = sys.stdin.read().strip()
+        if prompt_text:
+            return prompt_text
+
+    raise ValueError("prompt is required either as argv text or stdin")
+
+
+def get_tools(
+    runtime_environment: RuntimeEnvironment | None = None,
+    exec_mode: bool = False,
+):
+    from .tools import (
+        ApplyPatchTool,
+        CloseAgentTool,
+        CodeModeManager,
+        ExecTool,
+        ExecCommandTool,
+        GrepFilesTool,
+        ListDirTool,
+        ReadFileTool,
+        RequestPermissionsTool,
+        RequestUserInputTool,
+        ResumeAgentTool,
+        Registry,
+        SendInputTool,
+        ShellCommandTool,
+        ShellTool,
+        SpawnAgentTool,
+        UnifiedExecManager,
+        UpdatePlanTool,
+        ViewImageTool,
+        WaitAgentTool,
+        WaitTool,
+        WebSearchTool,
+        WriteStdinTool,
+    )
+
+    runtime_environment = runtime_environment or create_runtime_environment()
+    registry = Registry()
+    code_mode_manager = CodeModeManager(registry)
+    unified_exec_manager = UnifiedExecManager()
+    exec_tool = ExecTool(code_mode_manager)
+    wait_tool = WaitTool(code_mode_manager)
+    web_search_tool = WebSearchTool()
+    update_plan_tool = UpdatePlanTool(runtime_environment.plan_store)
+    request_user_input_tool = RequestUserInputTool(
+        runtime_environment.request_user_input_manager
+    )
+    request_permissions_tool = RequestPermissionsTool(
+        runtime_environment.request_permissions_manager
+    )
+    spawn_agent_tool = SpawnAgentTool(runtime_environment.subagent_manager)
+    send_input_tool = SendInputTool(runtime_environment.subagent_manager)
+    resume_agent_tool = ResumeAgentTool(runtime_environment.subagent_manager)
+    wait_agent_tool = WaitAgentTool(runtime_environment.subagent_manager)
+    close_agent_tool = CloseAgentTool(runtime_environment.subagent_manager)
+    apply_patch_tool = ApplyPatchTool()
+    shell_tool = ShellTool()
+    shell_command_tool = ShellCommandTool()
+    exec_command_tool = ExecCommandTool(unified_exec_manager)
+    write_stdin_tool = WriteStdinTool(unified_exec_manager)
+    grep_files_tool = GrepFilesTool()
+    read_file_tool = ReadFileTool()
+    list_dir_tool = ListDirTool()
+    view_image_tool = ViewImageTool()
+    if exec_mode:
+        registry.register(exec_command_tool)
+        registry.register(write_stdin_tool)
+        registry.register(update_plan_tool)
+        registry.register(request_user_input_tool)
+        registry.register(apply_patch_tool)
+        registry.register(web_search_tool)
+        registry.register(view_image_tool)
+        registry.register(spawn_agent_tool)
+        registry.register(send_input_tool)
+        registry.register(resume_agent_tool)
+        registry.register(wait_agent_tool)
+        registry.register(close_agent_tool)
+        return registry
+
+    registry.register(shell_tool)
+    registry.register(shell_command_tool)
+    registry.register(exec_command_tool)
+    registry.register(write_stdin_tool)
+    registry.register(exec_tool)
+    registry.register(wait_tool)
+    registry.register(web_search_tool)
+    registry.register(update_plan_tool)
+    registry.register(request_user_input_tool)
+    registry.register(request_permissions_tool)
+    registry.register(spawn_agent_tool)
+    registry.register(send_input_tool)
+    registry.register(resume_agent_tool)
+    registry.register(wait_agent_tool)
+    registry.register(close_agent_tool)
+    registry.register(apply_patch_tool)
+    registry.register(grep_files_tool)
+    registry.register(read_file_tool)
+    registry.register(list_dir_tool)
+    registry.register(view_image_tool)
+    return registry
+
+
+def get_subagent_tools(runtime_environment: RuntimeEnvironment | None = None):
+    from .tools import (
+        ApplyPatchTool,
+        ExecCommandTool,
+        Registry,
+        UnifiedExecManager,
+        UpdatePlanTool,
+        ViewImageTool,
+        WebSearchTool,
+        WriteStdinTool,
+    )
+
+    runtime_environment = runtime_environment or create_runtime_environment()
+    registry = Registry()
+    unified_exec_manager = UnifiedExecManager()
+    registry.register(ExecCommandTool(unified_exec_manager))
+    registry.register(WriteStdinTool(unified_exec_manager))
+    registry.register(UpdatePlanTool(runtime_environment.plan_store))
+    registry.register(ApplyPatchTool())
+    registry.register(WebSearchTool())
+    registry.register(ViewImageTool())
+    return registry
+
+
+def build_runtime(
+    config_path: str,
+    profile: str | None,
+    system_prompt: str | None,
+    client,
+    session_mode: CliSessionMode = "exec",
+    collaboration_mode: CollaborationMode = DEFAULT_COLLABORATION_MODE,
+) -> AgentRuntime:
+    use_tui_context = session_mode == "tui"
+    context_manager = ContextManager.from_codex_config(
+        config_path,
+        profile,
+        base_instructions_override=system_prompt,
+        collaboration_mode=collaboration_mode,
+        include_collaboration_instructions=use_tui_context,
+    )
+    subagent_context_manager = ContextManager.from_codex_config(
+        config_path,
+        profile,
+        base_instructions_override=system_prompt,
+        include_collaboration_instructions=False,
+    )
+    runtime_environment = create_runtime_environment()
+    runtime_environment.request_user_input_manager.set_handler(None)
+    runtime_environment.request_permissions_manager.set_handler(None)
+
+    def make_subagent_runtime_builder(base_client):
+        def build_subagent_runtime(
+            model_override: str | None,
+            reasoning_effort_override: str | None,
+            initial_history=(),
+            session_id: str | None = None,
+        ) -> AgentRuntime:
+            nested_client = base_client.with_overrides(
+                model_override,
+                reasoning_effort_override,
+                session_id=session_id,
+                openai_subagent="collab_spawn",
+            )
+            subagent_runtime_environment = create_runtime_environment()
+            subagent_runtime_environment.request_user_input_manager.set_handler(None)
+            subagent_runtime_environment.request_permissions_manager.set_handler(None)
+            subagent_runtime_environment.subagent_manager.set_runtime_builder(
+                make_subagent_runtime_builder(nested_client)
+            )
+            sub_agent = AgentLoop(
+                nested_client,
+                get_subagent_tools(subagent_runtime_environment),
+                subagent_context_manager,
+                initial_history=tuple(initial_history),
+            )
+            return AgentRuntime(
+                sub_agent, runtime_environment=subagent_runtime_environment
+            )
+
+        return build_subagent_runtime
+
+    runtime_environment.subagent_manager.set_runtime_builder(
+        make_subagent_runtime_builder(client)
+    )
+    return AgentRuntime(
+        AgentLoop(
+            client, get_tools(runtime_environment, exec_mode=True), context_manager
+        ),
+        runtime_environment=runtime_environment,
+    )
+
+
+def format_turn_output(result, json_mode: bool) -> str:
+    if json_mode:
+        return json.dumps(asdict(result), ensure_ascii=False, indent=2)
+    return result.output_text or ""
+
+
+def _build_model_client(
+    config_path: str,
+    profile: str | None,
+    timeout_seconds: float,
+    managed_responses_base_url: str | None = None,
+    vllm_endpoint: str | None = None,
+    use_chat_completion: bool = False,
+):
+    load_codex_dotenv(config_path)
+    provider_config = ResponsesProviderConfig.from_codex_config(
+        config_path,
+        profile,
+    )
+    url, key_env = provider_config.base_url, provider_config.api_key_env
+    if managed_responses_base_url is not None:
+        url, key_env = (
+            managed_responses_base_url,
+            LOCAL_RESPONSES_SERVER_API_KEY_ENV,
+        )
+        os.environ.setdefault(LOCAL_RESPONSES_SERVER_API_KEY_ENV, "dummy")
+    elif vllm_endpoint or use_chat_completion:
+        if vllm_endpoint:
+            managed_server = launch_chat_completion_compat_server(
+                vllm_endpoint,
+                model_provider="vllm",
+            )
+        else:
+            managed_server = launch_chat_completion_compat_server(
+                provider_config.base_url,
+                provider_config.api_key_env,
+                model_provider=provider_config.provider_name,
+            )
+        atexit.register(managed_server.stop)
+        url, key_env = (
+            managed_server.base_url,
+            LOCAL_RESPONSES_SERVER_API_KEY_ENV,
+        )
+        os.environ.setdefault(LOCAL_RESPONSES_SERVER_API_KEY_ENV, "dummy")
+
+    provider_config = replace(
+        provider_config,
+        base_url=url,
+        api_key_env=key_env,
+    )
+    return ResponsesModelClient(
+        provider_config,
+        timeout_seconds,
+        originator=CLI_ORIGINATOR,
+    )
+
+
+async def prompt_request_user_input(
+    view: CliSessionView,
+    payload: dict[str, object],
+) -> dict[str, object] | None:
+    view.finish_stream()
+    view.pause_spinner()
+    view.write_line("[request_user_input] waiting for user response")
+    answers: dict[str, dict[str, list[str]]] = {}
+    try:
+        for question in payload.get("questions", []):
+            if not isinstance(question, dict):
+                continue
+            header = str(question.get("header", "")).strip()
+            question_text = str(question.get("question", "")).strip()
+            question_id = str(question.get("id", "")).strip()
+            if header:
+                view.write_line(f"[{header}] {question_text}")
+            else:
+                view.write_line(question_text)
+
+            options = question.get("options") or []
+            if isinstance(options, list):
+                for index, option in enumerate(options, start=1):
+                    if not isinstance(option, dict):
+                        continue
+                    label = str(option.get("label", "")).strip()
+                    description = str(option.get("description", "")).strip()
+                    view.write_line(f"  {index}. {label} - {description}")
+            view.write_line("  0. Other")
+
+            try:
+                raw_answer = await view.prompt_async("answer> ")
+            except EOFError:
+                return None
+            answer_text = raw_answer.strip()
+            if not answer_text:
+                return None
+
+            selected_answer = answer_text
+            if answer_text.isdigit() and isinstance(options, list):
+                choice = int(answer_text)
+                if 1 <= choice <= len(options):
+                    option = options[choice - 1]
+                    if isinstance(option, dict):
+                        selected_answer = (
+                            str(option.get("label", "")).strip() or answer_text
+                        )
+                elif choice == 0:
+                    try:
+                        raw_answer = await view.prompt_async("other> ")
+                    except EOFError:
+                        return None
+                    selected_answer = raw_answer.strip()
+                    if not selected_answer:
+                        return None
+
+            answers[question_id] = {"answers": [selected_answer]}
+
+        return {"answers": answers}
+    finally:
+        view.resume_spinner()
+
+
+async def prompt_request_permissions(
+    view: CliSessionView,
+    payload: dict[str, object],
+) -> dict[str, object] | None:
+    view.finish_stream()
+    view.pause_spinner()
+    view.write_line("[request_permissions] user approval required")
+    reason = payload.get("reason")
+    if reason:
+        view.write_line(f"Reason: {reason}")
+    view.write_line("Requested permissions:")
+    view.write_line(
+        json.dumps(payload.get("permissions", {}), ensure_ascii=False, indent=2)
+    )
+    view.write_line("Choose: [n] deny / [t] grant for turn / [s] grant for session")
+    try:
+        raw_answer = await view.prompt_async("permissions> ")
+    except EOFError:
+        return None
+    finally:
+        view.resume_spinner()
+
+    answer = raw_answer.strip().lower()
+    if answer in {"t", "turn", "y", "yes"}:
+        return {
+            "permissions": payload.get("permissions", {}),
+            "scope": "turn",
+        }
+    if answer in {"s", "session"}:
+        return {
+            "permissions": payload.get("permissions", {}),
+            "scope": "session",
+        }
+    return {
+        "permissions": {},
+        "scope": "turn",
+    }
+
+
+async def run_interactive_session(
+    runtime: AgentRuntime,
+    json_mode: bool,
+) -> int:
+    worker = asyncio.create_task(runtime.run_forever())
+    view = CliSessionView()
+    model_client = runtime._agent_loop._model_client
+    runtime.set_event_handler(view.handle_event)
+    pending_turn_tasks: set[asyncio.Task[None]] = set()
+    runtime_environment = runtime.runtime_environment
+    if runtime_environment is None:
+        runtime_environment = create_runtime_environment()
+        runtime.runtime_environment = runtime_environment
+    runtime_environment.request_user_input_manager.set_handler(
+        lambda payload: prompt_request_user_input(view, payload)
+    )
+    runtime_environment.request_permissions_manager.set_handler(
+        lambda payload: prompt_request_permissions(view, payload)
+    )
+    view.write_line("pycodex interactive mode. Type /exit to quit.")
+    view.write_line("Extra commands: /history, /title, /model")
+    try:
+
+        def has_pending_turn_tasks() -> bool:
+            pending_turn_tasks.difference_update(
+                task for task in tuple(pending_turn_tasks) if task.done()
+            )
+            return bool(pending_turn_tasks)
+
+        async def wait_for_turn_result(future) -> None:
+            try:
+                result = await future
+            except Exception as exc:  # pragma: no cover - defensive surface
+                if str(exc) == "submission interrupted":
+                    return
+                view.show_error(str(exc))
+                return
+
+            if json_mode:
+                view.write_line(format_turn_output(result, True))
+
+        while True:
+            try:
+                raw_line = await view.poll_prompt("pycodex> ")
+            except EOFError:
+                break
+            if raw_line is None:
+                await asyncio.sleep(0.05)
+                continue
+
+            prompt_text = raw_line.strip()
+            if not prompt_text:
+                continue
+            if prompt_text in EXIT_COMMANDS:
+                break
+            if prompt_text == HISTORY_COMMAND:
+                view.show_history()
+                continue
+            if prompt_text == TITLE_COMMAND:
+                view.show_title()
+                continue
+            if prompt_text.startswith(f"{QUEUE_COMMAND} "):
+                queued_text = prompt_text[len(QUEUE_COMMAND) :].strip()
+                if not queued_text:
+                    view.write_line("Usage: /queue <message>")
+                    continue
+                try:
+                    submission_id, future = await runtime.enqueue_user_turn(
+                        queued_text, queue="enqueue"
+                    )
+                    view.show_steer_queued(submission_id, queued_text)
+                    turn_task = asyncio.create_task(wait_for_turn_result(future))
+                    pending_turn_tasks.add(turn_task)
+                except Exception as exc:  # pragma: no cover - defensive surface
+                    view.show_error(str(exc))
+                continue
+            if prompt_text == MODEL_COMMAND:
+                view.write_line(
+                    f"Current model: {getattr(model_client, 'model', None) or 'unavailable'}"
+                )
+                models = await model_client.list_models()
+                view.write_line(f"Available models: {', '.join(models)}")
+                continue
+            if prompt_text.startswith(f"{MODEL_COMMAND} "):
+                if has_pending_turn_tasks():
+                    view.write_line(
+                        "Cannot change model while work is running or queued in steer mode."
+                    )
+                    continue
+                model_name = prompt_text[len(MODEL_COMMAND) :].strip()
+                if not model_name:
+                    view.write_line("Usage: /model <model>")
+                    continue
+
+                model_client.model = model_name
+                view.write_line(f"Switched model to {model_name}.")
+                continue
+
+            try:
+                steered = has_pending_turn_tasks()
+                submission_id, future = await runtime.enqueue_user_turn(
+                    prompt_text,
+                    queue="steer",
+                )
+                if steered:
+                    view.schedule_steer_inserted(submission_id, prompt_text)
+                turn_task = asyncio.create_task(wait_for_turn_result(future))
+                pending_turn_tasks.add(turn_task)
+                continue
+            except Exception as exc:  # pragma: no cover - defensive surface
+                view.show_error(str(exc))
+                continue
+    finally:
+        runtime_environment.request_user_input_manager.set_handler(None)
+        runtime_environment.request_permissions_manager.set_handler(None)
+        await runtime.shutdown()
+        await worker
+        if pending_turn_tasks:
+            await asyncio.gather(*pending_turn_tasks, return_exceptions=True)
+        view.close()
+
+    return 0
+
+
+async def run_cli(args: argparse.Namespace) -> int:
+    runtime = None
+    worker = None
+    try:
+        if args.put is not None and args.call:
+            raise ValueError("--put and --call cannot be combined")
+        if args.put is not None and args.prompt:
+            raise ValueError("--put does not accept prompt text")
+        configure_loguru()
+        if args.put is not None:
+            def emit_put_log(message: str) -> None:
+                print(message, flush=True)
+
+            call_spec = upload_codex_home(args.put, event_handler=emit_put_log)
+            emit_put_log(f"[put] testing call: {call_spec}")
+            with tempfile.TemporaryDirectory(prefix="pycodex-put-call-test-") as tmpdir:
+                config_path = bootstrap_called_home(call_spec, storage_root=tmpdir)
+            emit_put_log(f"[put] call test ok: {config_path.name}")
+            print("[put] one-click start:", flush=True)
+            print(f"pycodex --call {shlex.quote(call_spec)}", flush=True)
+            return 0
+        if args.call:
+            config_path = bootstrap_called_home(args.call)
+            args.config = str(config_path)
+            os.environ["CODEX_HOME"] = str(config_path.parent)
+        client = _build_model_client(
+            args.config,
+            args.profile,
+            args.timeout_seconds,
+            vllm_endpoint=args.vllm_endpoint,
+            use_chat_completion=args.use_chat_completion,
+        )
+
+        runtime = build_runtime(
+            args.config,
+            args.profile,
+            args.system_prompt,
+            client,
+            session_mode="tui",
+        )
+        if should_run_interactive(args.prompt, sys.stdin.isatty()):
+            return await run_interactive_session(
+                runtime,
+                args.json,
+            )
+        else:
+            prompt_text = resolve_prompt_text(args.prompt)
+            worker = asyncio.create_task(runtime.run_forever())
+            result = await runtime.submit_user_turn(prompt_text)
+            print(format_turn_output(result, args.json))
+            return 0
+    except Exception as exc:
+        print(f"Error: {exc}", file=sys.stderr)
+        return 1
+    finally:
+        if runtime is not None and worker is not None:
+            await runtime.shutdown()
+            await worker
+
+
+def main(argv: Sequence[str] | None = None) -> int:
+    raw_args = list(argv) if argv is not None else None
+    if raw_args is None:
+        raw_args = sys.argv[1:]
+
+    if raw_args and raw_args[0] == "doctor":
+        from .doctor import build_doctor_parser, run_doctor_cli
+
+        parser = build_doctor_parser()
+        args = parser.parse_args(raw_args[1:])
+        try:
+            return asyncio.run(run_doctor_cli(args))
+        except ValueError as exc:
+            parser.error(str(exc))
+        except KeyboardInterrupt:
+            return 130
+        return 0
+
+    parser = build_parser()
+    args = parser.parse_args(raw_args)
+
+    try:
+        return asyncio.run(run_cli(args))
+    except ValueError as exc:
+        parser.error(str(exc))
+    except KeyboardInterrupt:
+        return 130
+    return 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())