MendCodeto code
Shape how AI codes: prompt modes, model roles, smart permissions, memory control, packages, worktrees, TSM, mflow, and a TUI your team can make its own.
A harness you can shape.
A workflow your team can share.
AI Coding Terminal
Open the public mendcode CLI in any repository, start an interactive coding TUI, or run a task with an initial message from the same local harness.
Prompt Modes
Choose minimal for low-noise experiments, focus for normal provider-family behavior, or full when package, memory, workflow, and runtime policy context should be present.
Model Roles
Assign different models to default, small, plan, build, code, review, subagent, title, compaction, summary, memoryExtractor, and permissionReviewer work.
Smart Permissions
Choose approval, smart, or full_access. Smart mode can route configured actions through a permissionReviewer role, and if that reviewer is missing MendCode asks instead of silently approving.
Approval-Gated Memory
Keep memory intentional. Generated memories become proposals first, project and global scopes stay separate, and direct saves require explicit user intent.
Custom TUI Profiles
Shape prompt chrome, lead strings, bottom status, home title, pet, chat presentation, themes, widgets, routes, dialogs, and plugins without patching runtime internals.
Package System
Bundle commands, agents, modes, skills, prompts, MCP config, TUI profiles, model roles, memory settings, permissions, and scripts into reusable team packages.
Agent Coordination
Use optional mflow coordination, local locks, TSM, and worktree orchestration when several agents or terminals need to work without stomping each other.
Usage Insights
See local usage signals like tokens, sessions, AI time, words, response load, top tools, agents, models, and daily activity from inside the TUI.
Install. Setup.
Then start coding.
Tune the
coding harness.
MendCode makes the AI coding terminal explicit: prompt mode controls how much harness context is added, focus profiles adapt behavior per provider family, and TUI profiles keep the interface shareable without patching runtime internals.
Customizable home screen
Know your usage.
Keep it local.
Inspect global, project, or directory-level evidence from inside the TUI: tokens, sessions, AI time, tools, agents, models, changed files, cache mix, and daily activity — without cloud analytics.
Remember with
review gates.
Memory Center is the TUI workspace for reviewing saved memories, pending proposals, categories, project scopes, and Dream maintenance without turning every chat into permanent context.
Package your
harness.
Turn a tuned MendCode environment into a reusable bundle: command palette, agents, modes, prompts, TUI profile, model policy, permissions, memory defaults, and worktree rules. No provider tokens or machine-local secrets.
Control is
non-negotiable.
MendCode keeps control explicit: choose approval, smart, or full_access; keep memory intentional; route smart reviews through a model role; and share packages without provider tokens or local secrets.
Permission modes
Choose approval, smart, or full_access depending on how much review you want in the current project.
No secrets in packages
Provider tokens, auth files, local credentials, and machine-local state are not part of shareable packages.
Intentional memory
Memory is opt-in. Generated memories become proposals first, and you decide what persists.
Risk-aware review
Smart permissions can route risky shell, script, and delete prompts through a permissionReviewer model role before execution.
The AI coding terminal
with runtime controls.
MendCode is not just a wrapper around a model. It is the public mendcode CLI, interactive coding TUI, prompt-mode controller, model-role runtime, package system, and local coordination harness for serious repo work.
Prompt modes
Minimal, focus, and full let you control harness context instead of hoping the prompt behaves.
Model roles
Route plan, build, review, subagent, summary, memory extraction, and permission review independently.
Developer harness
Packages, worktrees, TSM, and mflow turn the CLI into a repeatable team workflow.
Custom terminal
Tune prompt chrome, status, home identity, widgets, packages, and profiles from the TUI.
Open the AI coding terminal, run a task, inspect setup, or ask for repo status from the same CLI runtime.
# open the coding terminalmendcode# inspect provider, models, budget, prompt, package, TUI, memory, permissionsmendcode setup status# start with intentmendcode run "review this repo and draft a plan"# quick control-plane turnmendcode chat "summarize current status"
“Make the terminal configurable, not hardcoded.”
MendCode
Design principle, TUI
profile-driven terminal
Trusted by forward-thinking teams
Start the AI coding
terminal in your repo.
Install the public mendcode CLI, run setup once, then choose prompt mode, model roles, permissions, memory, packages, TUI profiles, usage insights, worktrees, TSM, and mflow around your workflow.
.-. (o o) /|[+]|\ /_|___|_\ \_/