# Copy this file to .env, then fill in only the providers you use.
# Do not commit .env or real tokens.

# Default model. Override per run with: aidd-intern --model <model-id>
AIDD_INTERN_DEFAULT_MODEL_ID=openrouter/openai/gpt-5.2
AIDD_INTERN_MODELS_CONFIG=configs/models.json

# LLM providers. Set at least one key that matches your selected model.
OPENAI_API_KEY=
ANTHROPIC_API_KEY=
OPENROUTER_API_KEY=
SILICONFLOW_API_KEY=
GEMINI_API_KEY=
DASHSCOPE_API_KEY=
DEEPSEEK_API_KEY=
KIMI_API_KEY=
MINIMAX_API_KEY=
XAI_API_KEY=
NVIDIA_API_KEY=
NOVITA_API_KEY=
TENCENT_API_KEY=
XIAOMI_API_KEY=
ARCEE_API_KEY=

# Real Google Search through Google Custom Search JSON API.
# Both values are required for the Google provider.
GOOGLE_SEARCH_API_KEY=
GOOGLE_SEARCH_ENGINE_ID=

# Optional aliases also recognized by the search tool.
GOOGLE_API_KEY=
GOOGLE_CSE_ID=

# Hugging Face and GitHub tools.
HF_TOKEN=
GITHUB_TOKEN=

# Local or LAN OpenAI-compatible inference servers.
LOCAL_LLM_BASE_URL=http://localhost:8000
LOCAL_LLM_API_KEY=
OLLAMA_BASE_URL=http://localhost:11434
VLLM_BASE_URL=http://localhost:8000
LMSTUDIO_BASE_URL=http://127.0.0.1:1234
LLAMACPP_BASE_URL=http://localhost:8080

# Optional local ProteinMCP setup.
AIDD_INTERN_PROTEINMCP_HOME=~/.cache/aidd-intern/proteinmcp
AIDD_INTERN_ENABLE_PROTEINMCP=0

# Optional update check controls.
# Set to 1 to suppress read-only version checks in doctor/interactive startup.
AIDD_INTERN_DISABLE_UPDATE_CHECK=0

# TencentDB-Agent-Memory 与 MemU 长期分层记忆配置项
# 用于存储和语义化检索智能体长短期记忆，大幅降低 Context 窗口开销
MEMU_API_KEY=
MEMU_BASE_URL=https://api.memu.so