MenteE SWE

An autonomous SWE agent in your terminal. It investigates your repository, plans, makes minimal edits, runs your tests, debugs failures, and reports back with evidence — never claiming success without proof.

Bring your own intelligence: MenteE is model-agnostic. Kimi, GLM, and more plug in as replaceable providers. The product is the harness: tools, verification loop, context management, and orchestration.

Install

npm install -g @menteeai/menteeswe

Quick start

# one-time configuration (choose Kimi or GLM, paste API key)
mentee config

# run a task in your project
cd your-project
mentee "fix the failing auth tests"

Or try without installing: npx @menteeai/menteeswe "your task".

Providers

| Provider | Default model | Env var | |---|---|---| | Kimi (Moonshot) | kimi-k2.7-code | MENTEE_KIMI_API_KEY | | GLM (Zhipu CN) | glm-4.6 | MENTEE_GLM_API_KEY | | Z.ai (GLM international) | glm-4.6 | MENTEE_ZAI_API_KEY | | mock (testing) | — | no key needed |

Select with -p kimi / -p glm / -p zai, override the model with -m <id>.

Inside the TUI

The interactive CLI stays open between tasks. Type a task and press Enter, or use slash commands:

/provider zai        switch provider (kimi, glm, zai, mock)
/model glm-4.5-air   set the model for the current provider
/key <api-key>       add the API key for the current provider (saved to ~/.mentee/config.json)
/help                list commands
/exit                quit

Safety model

• safe — read-only commands (git status, npm test, ls) run automatically
• restricted — installs, git history changes → interactive approval (y / a / n), or auto-approve with --yes
• dangerous — rm -rf, sudo, force pushes → always blocked

All file edits are constrained to the current workspace. Sessions log to ~/.mentee/logs/.

CLI

mentee [task...]                  run a task (interactive TUI in a terminal)
mentee config                     provider/key setup wizard
mentee -p glm "task"              choose provider
mentee -m kimi-k2-turbo-preview   override model
mentee --yes --no-tui "task"      headless mode for scripts/CI

Development

npm install
npm run dev        # run from source
npm test           # unit + integration tests (mock provider, no API keys needed)
npm run typecheck  # tsc --noEmit
npm run build      # bundle to dist/
npm run eval       # eval suite against fixture repos (mock provider)

Architecture

See ../MAIN_GOAL.md for the full product plan: phases, verification loop, subagents, context engineering, and routing roadmap.