hack/tools/scripts/llm-key-hunter/results/_cache/content/blaise-shyaka_expenses-log-ai-agent.json

{"repo": "blaise-shyaka/expenses-log-ai-agent", "files": {"expenses-agent/README.md": {"17870b609997b566da7cf90b8493ed1af49fb218": {"t": 1785746613.190167, "n": 2469, "h": "9179b6ddd8ece146", "c": "# Expenses Agent\n\nLangGraph AI agent that understands, categorizes, and records expenses. Talks to the user via a CopilotKit chat interface and calls tools exposed by the MCP server. Runs as a FastAPI server on port 8123.\n\n## Prerequisites\n\n- Python 3.13+\n- [uv](https://docs.astral.sh/uv/getting-started/installation/), a Python package manager\n- [Expenses MCP](../expenses-mcp/README.md) running at `http://localhost:8124`\n\n## 1. Configure Environment\n\n```bash\ncp .env.example .env\n```\n\nOpen `.env` and fill in your LLM credentials. You need **either** DeepSeek **or** Ollama. Not both. We chose deepseek, since it's good value for money.\n\n### Option A: DeepSeek (cloud)\n\nGet an API key at [platform.deepseek.com](https://platform.deepseek.com):\n\n```\nDEEPSEEK_API_KEY=sk-...\n```\n\n### Option B: Ollama (local)\n\nRun Ollama with llama3.2, or any other model you're comfortable with, then set its URL:\n\n```bash\ndocker run -d -v ollama:/root/.ollama -p 11434:11434 --name ollama ollama/ollama\ndocker exec -it ollama ollama pull llama3.2\n```\n\n```\nOLLAMA_URL=http://localhost:11434\n```\n\n> If `OLLAMA_URL` is set, it takes precedence and DeepSeek is ignored.\n\n### MCP Server URL\n\nThe default value in `.env.example` works as-is for local development:\n\n```\nMCP_URL=http://localhost:8124\n```\n\n> The agent connects to this URL on startup to load its tools. It retries with exponential backoff, so it tolerates the MCP server not being immediately ready.\n\n## 2. Install Dependencies\n\n```bash\nuv sync\n```\n\n## 3. Start the Agent\n\n### Option A: With LangGraph Studio (recommended for development)\n\n```bash\nuv run langgraph dev\n```\n\nThis starts the agent and prints three URLs:\n\n```\n🚀 API:       http://127.0.0.1:2024\n🎨 Studio UI: https://smith.langchain.com/studio/?baseUrl=http://127.0.0.1:2024\n📚 API Docs:  http://127.0.0.1:2024/docs\n```\n\nOpen the Studio URL to visually inspect the agent graph, step through tool calls, and replay conversations. The server supports hot-reloading. Code changes are reflected immediately.\n\n> **Safari users:** Safari blocks localhost connections. Use `uv run langgraph dev --tunnel` instead.\n\n> **Optional tracing:** To persist traces to LangSmith, add `LANGSMITH_API_KEY` to your `.env` (get one free at [smith.langchain.com/settings](https://smith.langchain.com/settings)).\n\n### Option B: Standalone server (no Studio)\n\n```bash\nuv run python main.py\n```\n\nStarts the agent at `http://localhost:8123`. Use this mode when running alongside the UI.\n"}, "2a0adf0cb5d719934d1737662bfc08c214bd1cfb": {"t": 1785766284.6552474, "n": 2469, "h": "9179b6ddd8ece146", "c": "# Expenses Agent\n\nLangGraph AI agent that understands, categorizes, and records expenses. Talks to the user via a CopilotKit chat interface and calls tools exposed by the MCP server. Runs as a FastAPI server on port 8123.\n\n## Prerequisites\n\n- Python 3.13+\n- [uv](https://docs.astral.sh/uv/getting-started/installation/), a Python package manager\n- [Expenses MCP](../expenses-mcp/README.md) running at `http://localhost:8124`\n\n## 1. Configure Environment\n\n```bash\ncp .env.example .env\n```\n\nOpen `.env` and fill in your LLM credentials. You need **either** DeepSeek **or** Ollama. Not both. We chose deepseek, since it's good value for money.\n\n### Option A: DeepSeek (cloud)\n\nGet an API key at [platform.deepseek.com](https://platform.deepseek.com):\n\n```\nDEEPSEEK_API_KEY=sk-...\n```\n\n### Option B: Ollama (local)\n\nRun Ollama with llama3.2, or any other model you're comfortable with, then set its URL:\n\n```bash\ndocker run -d -v ollama:/root/.ollama -p 11434:11434 --name ollama ollama/ollama\ndocker exec -it ollama ollama pull llama3.2\n```\n\n```\nOLLAMA_URL=http://localhost:11434\n```\n\n> If `OLLAMA_URL` is set, it takes precedence and DeepSeek is ignored.\n\n### MCP Server URL\n\nThe default value in `.env.example` works as-is for local development:\n\n```\nMCP_URL=http://localhost:8124\n```\n\n> The agent connects to this URL on startup to load its tools. It retries with exponential backoff, so it tolerates the MCP server not being immediately ready.\n\n## 2. Install Dependencies\n\n```bash\nuv sync\n```\n\n## 3. Start the Agent\n\n### Option A: With LangGraph Studio (recommended for development)\n\n```bash\nuv run langgraph dev\n```\n\nThis starts the agent and prints three URLs:\n\n```\n🚀 API:       http://127.0.0.1:2024\n🎨 Studio UI: https://smith.langchain.com/studio/?baseUrl=http://127.0.0.1:2024\n📚 API Docs:  http://127.0.0.1:2024/docs\n```\n\nOpen the Studio URL to visually inspect the agent graph, step through tool calls, and replay conversations. The server supports hot-reloading. Code changes are reflected immediately.\n\n> **Safari users:** Safari blocks localhost connections. Use `uv run langgraph dev --tunnel` instead.\n\n> **Optional tracing:** To persist traces to LangSmith, add `LANGSMITH_API_KEY` to your `.env` (get one free at [smith.langchain.com/settings](https://smith.langchain.com/settings)).\n\n### Option B: Standalone server (no Studio)\n\n```bash\nuv run python main.py\n```\n\nStarts the agent at `http://localhost:8123`. Use this mode when running alongside the UI.\n"}, "5eae56ec8e3cd503290e64d07b95d203622b34eb": {"t": 1785766286.3682775, "n": 2364, "h": "78bf8efadae3a341", "c": "# Expenses Agent\n\nLangGraph AI agent that understands, categorizes, and records expenses. Communicates with the Expenses API to record and query data. Runs as a FastAPI server on port 8123.\n\n## Prerequisites\n\n- Python 3.13+\n- [uv](https://docs.astral.sh/uv/getting-started/installation/), a Python package manager\n- [Expenses API](../expenses-api/README.md) running at `http://localhost:8000`\n\n## 1. Configure Environment\n\n```bash\ncp .env.example .env\n```\n\nOpen `.env` and fill in your LLM credentials. You need **either** DeepSeek **or** Ollama. Not both. We chose deepseek, since it's good value for money.\n\n### Option A: DeepSeek (cloud)\n\nGet an API key at [platform.deepseek.com](https://platform.deepseek.com):\n\n```\nDEEPSEEK_API_KEY=sk-...\n```\n\n### Option B: Ollama (local)\n\nRun Ollama with llama3.2, or any other model you're comfortable with, then set its URL:\n\n```bash\ndocker run -d -v ollama:/root/.ollama -p 11434:11434 --name ollama ollama/ollama\ndocker exec -it ollama ollama pull llama3.2\n```\n\n```\nOLLAMA_URL=http://localhost:11434\n```\n\n> If `OLLAMA_URL` is set, it takes precedence and Deepseek is ignored.\n\n### Expenses API URL\n\nThe default value in `.env.example` works as-is for local development:\n\n```\nEXPENSES_API_URL=http://localhost:8000/api/v1\n```\n\n> `/api/v1` is the API's base path. All routes are prefixed with it.\n\n## 2. Install Dependencies\n\n```bash\nuv sync\n```\n\n## 3. Start the Agent\n\n### Option A: With LangGraph Studio (recommended for development)\n\n```bash\nuv run langgraph dev\n```\n\nThis starts the agent and prints three URLs:\n\n```\n🚀 API:       http://127.0.0.1:2024\n🎨 Studio UI: https://smith.langchain.com/studio/?baseUrl=http://127.0.0.1:2024\n📚 API Docs:  http://127.0.0.1:2024/docs\n```\n\nOpen the Studio URL to visually inspect the agent graph, step through tool calls, and replay conversations. The server supports hot-reloading. Code changes are reflected immediately.\n\n> **Safari users:** Safari blocks localhost connections. Use `uv run langgraph dev --tunnel` instead.\n\n> **Optional tracing:** To persist traces to LangSmith, add `LANGSMITH_API_KEY` to your `.env` (get one free at [smith.langchain.com/settings](https://smith.langchain.com/settings)).\n\n### Option B: Standalone server (no Studio)\n\n```bash\nuv run python main.py\n```\n\nStarts the agent at `http://localhost:8123`. Use this mode when running alongside the UI.\n"}}}}