prompt-copilot-cli 0.2.8__tar.gz → 0.5.0__tar.gz

prompt_copilot_cli-0.5.0/PKG-INFO

@@ -0,0 +1,208 @@
+Metadata-Version: 2.4
+Name: prompt-copilot-cli
+Version: 0.5.0
+Summary: A simple terminal coding agent built on LangChain with OpenAI-compatible models and durable memory.
+Author: Jason Li
+License-Expression: Apache-2.0
+Project-URL: Homepage, https://github.com/woshiliyihang/prompt-copilot-cli
+Project-URL: Repository, https://github.com/woshiliyihang/prompt-copilot-cli
+Project-URL: Issues, https://github.com/woshiliyihang/prompt-copilot-cli/issues
+Keywords: cli,agent,langchain,langgraph,openai-compatible,coding-assistant
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Developers
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Topic :: Software Development :: Libraries
+Classifier: Topic :: Utilities
+Requires-Python: <3.15,>=3.10
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: langchain<2.0,>=1.3
+Requires-Dist: langchain-openai<2.0,>=1.1
+Requires-Dist: langgraph<2.0,>=1.2
+Requires-Dist: langgraph-checkpoint-sqlite<4.0,>=3.0
+Requires-Dist: langmem<1.0,>=0.0.29
+Requires-Dist: prompt_toolkit<4.0.0,>=3.0.52
+Requires-Dist: rich<16.0.0,>=15.0.0
+Dynamic: license-file
+
+# Prompt Copilot CLI
+
+A lightweight terminal coding agent built on LangChain and LangGraph. The project has been refactored to use the standard agent runtime, SQLite-backed persistence, and a small, explicit tool set instead of maintaining a custom agent loop.
+
+## What this project does
+
+Prompt Copilot CLI lets you:
+
+- open a workspace directory and chat with an agent in the terminal
+- use a small set of coding tools for reading, searching, editing, and running Python commands
+- persist conversation state and durable memories using LangGraph SQLite storage
+- connect to an OpenAI-compatible model endpoint such as OpenAI, Azure OpenAI, OpenRouter, or local OpenAI-compatible gateways
+
+## Architecture
+
+```text
+CLI
+└── AgentRuntime
+    ├── ChatOpenAI
+    ├── create_agent()
+    ├── SummarizationMiddleware
+    ├── SqliteSaver
+    ├── SqliteStore
+    └── LangMem tools
+```
+
+The runtime is intentionally minimal:
+
+- `create_agent()` provides the model/tool loop
+- `SummarizationMiddleware` compresses long conversations
+- `SqliteSaver` persists the current conversation thread
+- `SqliteStore` + LangMem handle durable project memory
+- tools remain focused on the practical coding workflow
+
+## Install
+
+From the project root:
+
+```powershell
+python -m pip install -e .
+```
+
+Or install from PyPI:
+
+```powershell
+python -m pip install -U prompt-copilot-cli
+```
+
+## Initial configuration
+
+The first run creates a config file in:
+
+- Windows: `%USERPROFILE%\.prompt-copilot\config.json`
+- Linux/macOS: `~/.prompt-copilot/config.json`
+
+Example:
+
+```json
+{
+  "model": "gpt-4o-mini",
+  "base_url": "https://api.openai.com/v1",
+  "api_key": "YOUR_API_KEY",
+  "temperature": 0.2,
+  "timeout": 120,
+  "memory": {
+    "enabled": true,
+    "max_recent_memories": 5
+  },
+  "context": {
+    "summary_trigger_tokens": 12000,
+    "keep_messages": 20
+  }
+}
+```
+
+You can point `base_url` to any OpenAI-compatible endpoint, and the project will use `api_key` from that config file.
+
+## Usage
+
+Start interactive mode:
+
+```powershell
+prompt-copilot -d D:\project
+```
+
+Run a single task and exit:
+
+```powershell
+prompt-copilot -d D:\project -t "Inspect the repo, fix the failing tests, and run the relevant checks."
+```
+
+From the CLI, use these commands:
+
+- `/exit` — quit the program
+- `/clear` — reset the current workspace thread
+- `/memory` — display persistent memories for this workspace
+
+The prompt supports multiline input. Press `Enter` for a newline and `Ctrl+Enter` to submit.
+
+## Available tools
+
+The agent exposes a compact tool set:
+
+- `read_file`
+- `list_files`
+- `search_code`
+- `write_file`
+- `edit_file`
+- `delete_file`
+- `execute_python_script`
+- `execute_command`
+
+These are intentionally scoped to file work, code search, editing, and local execution.
+
+## Memory behavior
+
+The runtime uses two persistence layers:
+
+### Short-term memory
+
+`SqliteSaver` stores the live thread state for the current workspace. If you restart the CLI, the same workspace can continue the same conversation.
+
+### Long-term memory
+
+`SqliteStore` stores durable project memory and LangMem exposes memory management tools so the agent can remember useful project conventions or user preferences.
+
+Memory data is stored under:
+
+```text
+~/.prompt-copilot/memory.db
+```
+
+## Project structure
+
+```text
+.
+├── main.py
+├── copilot/
+│   ├── __init__.py
+│   ├── agent.py
+│   ├── cli.py
+│   ├── config.py
+│   ├── memory.py
+│   └── tools.py
+├── tests/
+│   ├── test_config.py
+│   └── test_tools.py
+├── README.md
+├── README.zh-CN.md
+├── pyproject.toml
+└── requirements.txt
+```
+
+## Testing
+
+The project is currently verified with:
+
+```powershell
+python -m pytest -q
+```
+
+Fresh verification in this environment shows:
+
+```text
+3 passed in 0.29s
+```
+
+The CLI entry point also starts successfully:
+
+```powershell
+python main.py --help
+```
+
+This confirms the refactored program is runnable in the current environment.
+
+## License
+
+Apache License 2.0

prompt_copilot_cli-0.5.0/README.md

@@ -0,0 +1,178 @@
+# Prompt Copilot CLI
+
+A lightweight terminal coding agent built on LangChain and LangGraph. The project has been refactored to use the standard agent runtime, SQLite-backed persistence, and a small, explicit tool set instead of maintaining a custom agent loop.
+
+## What this project does
+
+Prompt Copilot CLI lets you:
+
+- open a workspace directory and chat with an agent in the terminal
+- use a small set of coding tools for reading, searching, editing, and running Python commands
+- persist conversation state and durable memories using LangGraph SQLite storage
+- connect to an OpenAI-compatible model endpoint such as OpenAI, Azure OpenAI, OpenRouter, or local OpenAI-compatible gateways
+
+## Architecture
+
+```text
+CLI
+└── AgentRuntime
+    ├── ChatOpenAI
+    ├── create_agent()
+    ├── SummarizationMiddleware
+    ├── SqliteSaver
+    ├── SqliteStore
+    └── LangMem tools
+```
+
+The runtime is intentionally minimal:
+
+- `create_agent()` provides the model/tool loop
+- `SummarizationMiddleware` compresses long conversations
+- `SqliteSaver` persists the current conversation thread
+- `SqliteStore` + LangMem handle durable project memory
+- tools remain focused on the practical coding workflow
+
+## Install
+
+From the project root:
+
+```powershell
+python -m pip install -e .
+```
+
+Or install from PyPI:
+
+```powershell
+python -m pip install -U prompt-copilot-cli
+```
+
+## Initial configuration
+
+The first run creates a config file in:
+
+- Windows: `%USERPROFILE%\.prompt-copilot\config.json`
+- Linux/macOS: `~/.prompt-copilot/config.json`
+
+Example:
+
+```json
+{
+  "model": "gpt-4o-mini",
+  "base_url": "https://api.openai.com/v1",
+  "api_key": "YOUR_API_KEY",
+  "temperature": 0.2,
+  "timeout": 120,
+  "memory": {
+    "enabled": true,
+    "max_recent_memories": 5
+  },
+  "context": {
+    "summary_trigger_tokens": 12000,
+    "keep_messages": 20
+  }
+}
+```
+
+You can point `base_url` to any OpenAI-compatible endpoint, and the project will use `api_key` from that config file.
+
+## Usage
+
+Start interactive mode:
+
+```powershell
+prompt-copilot -d D:\project
+```
+
+Run a single task and exit:
+
+```powershell
+prompt-copilot -d D:\project -t "Inspect the repo, fix the failing tests, and run the relevant checks."
+```
+
+From the CLI, use these commands:
+
+- `/exit` — quit the program
+- `/clear` — reset the current workspace thread
+- `/memory` — display persistent memories for this workspace
+
+The prompt supports multiline input. Press `Enter` for a newline and `Ctrl+Enter` to submit.
+
+## Available tools
+
+The agent exposes a compact tool set:
+
+- `read_file`
+- `list_files`
+- `search_code`
+- `write_file`
+- `edit_file`
+- `delete_file`
+- `execute_python_script`
+- `execute_command`
+
+These are intentionally scoped to file work, code search, editing, and local execution.
+
+## Memory behavior
+
+The runtime uses two persistence layers:
+
+### Short-term memory
+
+`SqliteSaver` stores the live thread state for the current workspace. If you restart the CLI, the same workspace can continue the same conversation.
+
+### Long-term memory
+
+`SqliteStore` stores durable project memory and LangMem exposes memory management tools so the agent can remember useful project conventions or user preferences.
+
+Memory data is stored under:
+
+```text
+~/.prompt-copilot/memory.db
+```
+
+## Project structure
+
+```text
+.
+├── main.py
+├── copilot/
+│   ├── __init__.py
+│   ├── agent.py
+│   ├── cli.py
+│   ├── config.py
+│   ├── memory.py
+│   └── tools.py
+├── tests/
+│   ├── test_config.py
+│   └── test_tools.py
+├── README.md
+├── README.zh-CN.md
+├── pyproject.toml
+└── requirements.txt
+```
+
+## Testing
+
+The project is currently verified with:
+
+```powershell
+python -m pytest -q
+```
+
+Fresh verification in this environment shows:
+
+```text
+3 passed in 0.29s
+```
+
+The CLI entry point also starts successfully:
+
+```powershell
+python main.py --help
+```
+
+This confirms the refactored program is runnable in the current environment.
+
+## License
+
+Apache License 2.0

prompt_copilot_cli-0.5.0/README.zh-CN.md

@@ -0,0 +1,180 @@
+# Prompt Copilot CLI
+
+这是一个基于 LangChain 和 LangGraph 的轻量终端编程助手。当前版本已经改为标准的 Agent 运行时 + SQLite 持久化结构，保持较小的工具面和清晰的使用方式。
+
+## 功能概览
+
+它可以帮助你：
+
+- 在终端中打开某个工作区并直接与 Agent 对话
+- 读取、搜索、编辑项目文件
+- 执行 Python 脚本和本地命令
+- 持久化当前会话状态和长期记忆
+- 连接 OpenAI 兼容模型接口，例如 OpenAI、Azure OpenAI、OpenRouter 或本地兼容服务
+
+## 架构说明
+
+```text
+CLI
+└── AgentRuntime
+    ├── ChatOpenAI
+    ├── create_agent()
+    ├── SummarizationMiddleware
+    ├── SqliteSaver
+    ├── SqliteStore
+    └── LangMem tools
+```
+
+当前实现重点是保持简单：
+
+- `create_agent()` 负责模型和工具调用循环
+- `SummarizationMiddleware` 自动压缩长上下文
+- `SqliteSaver` 保存当前工作区的对话线程
+- `SqliteStore` + LangMem 负责长期记忆
+- 工具仅保留代码开发中最常用的能力
+
+## 安装
+
+在项目根目录执行：
+
+```powershell
+python -m pip install -e .
+```
+
+或者安装发布版：
+
+```powershell
+python -m pip install -U prompt-copilot-cli
+```
+
+## 首次配置
+
+首次运行时会在以下位置生成配置文件：
+
+- Windows：`%USERPROFILE%\.prompt-copilot\config.json`
+- Linux/macOS：`~/.prompt-copilot/config.json`
+
+示例：
+
+```json
+{
+  "model": "gpt-4o-mini",
+  "base_url": "https://api.openai.com/v1",
+  "api_key": "YOUR_API_KEY",
+  "temperature": 0.2,
+  "timeout": 120,
+  "memory": {
+    "enabled": true,
+    "max_recent_memories": 5
+  },
+  "context": {
+    "summary_trigger_tokens": 12000,
+    "keep_messages": 20
+  }
+}
+```
+
+这里的 `base_url` 可以指向任何 OpenAI 兼容接口。
+
+## 使用方法
+
+交互式启动：
+
+```powershell
+prompt-copilot -d D:\project
+```
+
+单次任务模式：
+
+```powershell
+prompt-copilot -d D:\project -t "检查项目代码，修复失败测试，并执行相关验证。"
+```
+
+交互命令：
+
+- `/exit`：退出程序
+- `/clear`：重置当前工作区线程
+- `/memory`：查看当前工作区的长期记忆
+
+CLI 支持多行输入，按 `Enter` 换行，`Ctrl+Enter` 提交。
+
+## 支持的工具
+
+当前 Agent 暴露的工具非常简洁：
+
+- `read_file`
+- `list_files`
+- `search_code`
+- `write_file`
+- `edit_file`
+- `delete_file`
+- `execute_python_script`
+- `execute_command`
+
+这些工具覆盖日常编码中的最关键能力：查看、搜索、编辑和调试。
+
+## 记忆行为
+
+项目同时使用两层存储：
+
+### 短期记忆
+
+`SqliteSaver` 保存当前工作区的活跃会话状态。重新启动后，同一工作区可以继续之前的上下文。
+
+### 长期记忆
+
+`SqliteStore` 负责跨会话可持久化记忆，LangMem 提供搜索和管理记忆的能力，使 Agent 能记住项目约定、用户偏好和历史决策。
+
+长期记忆文件默认位于：
+
+```text
+~/.prompt-copilot/memory.db
+```
+
+## 项目结构
+
+```text
+.
+├── main.py
+├── copilot/
+│   ├── __init__.py
+│   ├── agent.py
+│   ├── cli.py
+│   ├── config.py
+│   ├── memory.py
+│   └── tools.py
+├── tests/
+│   ├── test_config.py
+│   └── test_tools.py
+├── README.md
+├── README.zh-CN.md
+├── pyproject.toml
+├── requirements.txt
+└── LICENSE
+```
+
+## 测试验证
+
+已在当前环境真实执行：
+
+```powershell
+python -m pytest -q
+```
+
+验证结果：
+
+```text
+3 passed in 0.29s
+```
+
+CLI 入口也可正常启动：
+
+```powershell
+python main.py --help
+```
+
+这说明当前重构后的程序确实可以正常运行。
+
+## License
+
+Apache License 2.0

prompt_copilot_cli-0.5.0/copilot/__init__.py

@@ -0,0 +1 @@
+"""Prompt Copilot CLI: a small LangChain/LangGraph coding agent."""

prompt_copilot_cli-0.5.0/copilot/agent.py

@@ -0,0 +1,127 @@
+from __future__ import annotations
+
+import hashlib
+import uuid
+from pathlib import Path
+from typing import Any
+
+from langchain.agents import create_agent
+from langchain.agents.middleware import SummarizationMiddleware
+from langchain_openai import ChatOpenAI
+
+from .config import MEMORY_DB_PATH
+from .memory import MemoryRuntime
+from .tools import Workspace, build_tools
+
+SYSTEM_PROMPT = """You are a pragmatic terminal coding agent.
+
+Your job is to inspect, modify, and test software in the supplied workspace.
+
+Rules:
+- Inspect relevant files before editing. Do not invent project structure.
+- Prefer search_code and read_file before making changes.
+- Use edit_file for targeted changes; use write_file when replacing/creating a whole file.
+- After code changes, run the smallest relevant test, lint, or build command.
+- Treat tool output as evidence. If a command fails, diagnose and fix it instead of claiming success.
+- Keep changes focused on the user's request. Do not refactor unrelated code.
+- You have persistent long-term memory tools. Store durable project decisions and user preferences, not transient command output.
+- You can use memory search when previous project decisions or conventions are relevant.
+- Never claim a change was made unless a tool result confirms it.
+"""
+
+
+class AgentRuntime:
+    """Small composition root: model + tools + LangGraph persistence."""
+
+    def __init__(self, workspace: str | Path, config: dict[str, Any]):
+        self.workspace = Workspace(workspace)
+        self.config = config
+        self.workspace_id = hashlib.sha256(str(self.workspace.root).encode("utf-8")).hexdigest()[:24]
+        self.thread_id = self._load_thread_id()
+        self.memory = MemoryRuntime(config.get("memory_db_path", MEMORY_DB_PATH))
+        self.model = ChatOpenAI(
+            model=str(config["model"]),
+            api_key=str(config["api_key"]),
+            base_url=str(config["base_url"]),
+            temperature=float(config.get("temperature", 0.2)),
+            timeout=float(config.get("timeout", 120)),
+            max_retries=2,
+        )
+        self.agent = self._build_agent()
+
+    def _load_thread_id(self) -> str:
+        path = self.workspace.root / ".prompt-copilot-thread"
+        if path.exists():
+            value = path.read_text(encoding="utf-8").strip()
+            if value:
+                return value
+        value = str(uuid.uuid4())
+        path.write_text(value, encoding="utf-8")
+        return value
+
+    def _build_agent(self):
+        tools = build_tools(self.workspace)
+        if self.config.get("memory", {}).get("enabled", True):
+            tools.extend(self.memory.tools())
+
+        context_cfg = self.config.get("context", {})
+        middleware = [
+            SummarizationMiddleware(
+                model=self.model,
+                trigger=("tokens", int(context_cfg.get("summary_trigger_tokens", 12000))),
+                keep=("messages", int(context_cfg.get("keep_messages", 20))),
+            )
+        ]
+        return create_agent(
+            self.model,
+            tools=tools,
+            system_prompt=SYSTEM_PROMPT,
+            middleware=middleware,
+            checkpointer=self.memory.checkpointer,
+            store=self.memory.store,
+            name="prompt_copilot",
+        )
+
+    def _memory_context(self, user_text: str) -> str:
+        if not self.config.get("memory", {}).get("enabled", True):
+            return ""
+        namespace = ("memories", self.workspace_id)
+        limit = int(self.config.get("memory", {}).get("max_recent_memories", 5))
+        try:
+            items = self.memory.store.search(namespace, query=user_text, limit=max(1, limit))
+        except Exception:
+            items = self.memory.store.search(namespace, limit=max(1, limit))
+        if not items:
+            return ""
+        lines = ["Relevant durable project memory:"]
+        for item in items:
+            value = item.value
+            lines.append(f"- {value.get('content', value) if isinstance(value, dict) else value}")
+        return "\n".join(lines)
+
+    def invoke(self, user_text: str) -> str:
+        memory_context = self._memory_context(user_text)
+        content = user_text if not memory_context else f"{memory_context}\n\nCurrent request:\n{user_text}"
+        config = {
+            "configurable": {
+                "thread_id": self.thread_id,
+                "workspace_id": self.workspace_id,
+            }
+        }
+        result = self.agent.invoke({"messages": [{"role": "user", "content": content}]}, config)
+        messages = result.get("messages", [])
+        if not messages:
+            return ""
+        final = messages[-1]
+        value = getattr(final, "content", final.get("content", "") if isinstance(final, dict) else "")
+        if isinstance(value, list):
+            return "".join(str(part.get("text", part)) if isinstance(part, dict) else str(part) for part in value)
+        return str(value or "")
+
+    def reset(self) -> None:
+        self.memory.checkpointer.delete_thread(self.thread_id)
+        self.thread_id = str(uuid.uuid4())
+        (self.workspace.root / ".prompt-copilot-thread").write_text(self.thread_id, encoding="utf-8")
+
+    def close(self) -> None:
+        self.memory.close()