whet 0.1.0

An open-source terminal coding agent. Powered by local or cloud LLMs.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34

# whet configuration
# Copy to ~/.whet/config.toml

[llm]
provider = "ollama"          # "ollama" or "openai_compat"
model = "qwen3:8b"
base_url = "http://localhost:11434"
# api_key = "sk-..."         # Optional: required for some OpenAI-compatible servers
# streaming = true           # Enable streaming responses (default: false)

# OpenAI-compatible server examples:
# [llm]
# provider = "openai_compat"
# model = "gpt-3.5-turbo"
# base_url = "http://localhost:8080"   # llama.cpp server
# api_key = "sk-..."                   # Optional
# streaming = true

[agent]
max_iterations = 10
# permission_mode = "default"   # "default", "accept_edits", or "yolo"
# - default: Ask before file writes and command execution
# - accept_edits: Auto-approve file edits, ask for shell/git
# - yolo: No confirmation needed
# web_enabled = false            # Enable web_fetch and web_search tools (requires internet)

[memory]
database_path = "~/.whet/memory.db"

# MCP (Model Context Protocol) servers
# [[mcp.servers]]
# name = "filesystem"
# command = "npx"
# args = ["-y", "@modelcontextprotocol/server-filesystem", "/tmp"]