npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2026 – Pkg Stats / Ryan Hefner

monkeyscode-cli

v1.2.9

Published

MonkeysCode CLI — standalone terminal coding agent. Works independently of the IDE and Agent Manager.

Readme

MonkeysCode CLI

AI-powered coding agent for the terminal. Works on macOS, Linux, and Windows.

npm version GitHub Actions License

Quick Install

# macOS / Linux (installs `mc` to /usr/local/bin; override with MC_INSTALL_DIR)
curl -fsSL https://monkeyscode.com/cli/install.sh | bash

# Windows (PowerShell)
irm https://monkeyscode.com/cli/install.ps1 | iex

# npm (Node 20+)
npm install -g monkeyscode-cli

The binary is installed as mc, with monkeyscode (and mcode from the install script) as aliases. A Homebrew formula template lives in packaging/homebrew/, but there is no published tap yet.

Usage

Core Skills onboarding

mc plugin core-skills                 # stage MonkeysCode/core-skills from the signed catalog
mc plugin bootstrap                   # alias for core-skills
mc plugin enable core-skills          # review capabilities, confirm version trust, then enable
mc plugin core-skills --enable        # explicitly combine installation and normal enable review

Installation alone displays the capability report but does not enable skills or change existing profiles. --enable, --scope user|project and --yes use the existing plugin trust policy; --yes never bypasses managed approval. This is the same native pipeline as mc plugin install MonkeysCode/core-skills, for catalog-managed updates, not another injection of shared legacy app content. The packaged 24-definition library is declarative application content, not a signed plugin or a trust root. Catalog/API unavailability or verification failure fails closed: no unsigned local, packaged or git fallback. No automatic onboarding runs at CLI startup.

Public plugin discovery

mc plugin search [query] [--category <category>] [--app <app>] [--json]
mc plugin info <author/name>
mc plugin outdated [--json]
mc plugin update <name>
mc plugin update --all
# Interactive: /plugins browse [query] [--category <category>] [--app <app>]

Discovery uses https://marketplace.monkeyscode.com by default; set MONKEYSCODE_CATALOG_ORIGIN to another public HTTPS origin, or to an empty value to turn discovery off. It sends no account credentials. Metadata is not package trust. Signing pins are provisioned independently with MONKEYSCODE_PLUGIN_SIGNING_KEYS; never accept keys learned from discovery as trust roots. Bulk updates do not treat --yes as approval for expanded capabilities; each expansion needs review. Managed-policy pending trust must be resolved before enabling.

# Interactive REPL (optionally with a first prompt)
mc
mc "refactor the auth module"

# Headless (CI/CD)
mc -p "fix all lint errors" --headless --auto-approve

# Background job with auto-PR
mc -p "write tests" --background --auto-branch --auto-pr

# Pipe mode: stdin becomes the prompt, the answer goes to stdout
echo "explain this error" | mc

# Model selection
mc --model opus "deep architecture review"

Features

🤖 Multi-Model Support

Switch between MonkeysCode's own Capuchin model (the default), Claude Sonnet and Opus, Gemini 2.5 Pro, and GPT-4.1. mc models list shows the full list with pricing.

mc --model auto "fix the bug"   # Your configured default model
mc --model opus "review PR"     # Aliases: sonnet, opus, capuchin, or a full model id

📁 Full Tool Suite

The agent can edit files, run terminal commands, search code, and manage git — all locally.

Safe editing: When an old_string has only minor whitespace differences (tabs vs spaces, trailing space), the edit engine recovers it instead of rejecting. Identical old_string/new_string pairs are treated as no-ops and skipped — they never abort an otherwise valid batch. Preview and execution use the same engine, so what you see is what gets written.

Cross-platform search: grep_search prefers ripgrep, falls back to grep, and finally uses a built-in shell-free Node search when neither is available (e.g. a minimal Windows install). Unsupported regex semantics produce an explicit error, never a false "No matches".

🔒 Sandboxed Execution

Run agent commands in isolated sandboxes (bubblewrap on Linux, sandbox-exec on macOS, Docker anywhere). The best available provider is picked automatically; without one, --sandbox reports it as unavailable instead of pretending.

On Windows, GUI Windows Sandbox and AppContainer are not treated as verified headless isolation — the provider refuses to activate and directs you to Docker. Ordinary native Windows execution (without --sandbox) works without Docker.

mc --sandbox "run the migration script"
mc --sandbox --sandbox-network none "process sensitive data"   # none | local | full
mc --sandbox --sandbox-provider docker "try the install script"

🌐 Background Jobs

Submit long-running tasks to the cloud and get notified when done.

mc -p "write comprehensive tests" --background --auto-pr
mc jobs list
mc jobs logs abc123 --follow

🧠 Memory & Conventions

The agent learns your project's conventions and remembers them.

mc memory add "Always use camelCase for variables"
mc memory show
mc dream           # Consolidate session history into memory

⚡ CI/CD Integration

Run headlessly in GitHub Actions, GitLab CI, or any CI system. The action lives in this directory (apps/cli/action.yml), not at the repository root:

# .github/workflows/monkeyscode-mention.yml — answers "@monkeyscode …" comments, read-only
- uses: MonkeysCloud/monkeyscode/apps/cli@main
  with:
    api-key: ${{ secrets.MONKEYSCODE_API_KEY }}
    max-cost: '1'

Ready-made workflows are in workflows/: PR review (SARIF), @monkeyscode mentions, fix failing CI → PR, issue → draft PR, and nightly dependency updates. Each one limits what the agent may run, never lets it push, and skips fork runs that would expose your key. Give CI a restricted key (spend cap + model allowlist); the action warns when it isn't. See docs/cli/ci.md.

Anywhere else, call the CLI directly:

mc -p "fix the failing test" --allowed-tools "run_command(npm test:*)" "edit_file(src/**)" -o json --max-cost 2
git diff | mc -p "review this diff"            # stdin is appended to the prompt
mc -p -o stream-json "refactor utils"           # one JSON event per line

The exit code tells you what happened: 0 ok, 2 bad flags, 3 auth, 4 quota, 5 --max-cost/--max-tokens, 6 only denials, 124 timeout. Flags, the event stream, multi-turn stdin and permission prompts are in docs/cli/headless.md.

🔌 MCP Servers

Extend the agent with custom tools via the Model Context Protocol.

mc mcp browse                                   # 35+ curated servers
mc mcp add jira                                 # Add from the registry by id
mc mcp add my-tools npx -y my-mcp-server        # Or any command
mc mcp list

Servers start in the background when a session opens; their tools reach the model as mcp__<server>__<tool> (e.g. mcp__jira__create_issue) and follow permission rules like any other tool (mcp__jira__* allows or denies a whole server). /mcp shows status and tool counts and can reconnect, disable or enable a server for the session.

A server defined by the repository (.monkeyscode/settings.json, .monkeyscode/mcp.json) runs a command someone else chose, so it does not start until you approve it with /mcp approve <name> or mc mcp approve <name>. Editing the entry voids the approval. Servers you add with mc mcp add are approved automatically.

🧩 Commands, Subagents, Skills, Hooks & Plugins

Extend the agent with plain files in .monkeyscode/, in Claude Code's formats:

mc import claude --dry-run          # bring over .claude/ (commands, agents, skills, hooks, MCP)
mc import cursor                    # .cursor/rules, .cursorrules, .cursor/mcp.json
mc plugin install https://github.com/<owner>/<repo>   # guarded: nothing from it runs
mc plugin enable <name>             # review what it can do, then trust + enable
  • .monkeyscode/commands/*.md → /name ($ARGUMENTS, @file, !`cmd`, allowed-tools)
  • .monkeyscode/agents/*.md → subagents the model delegates to
  • .monkeyscode/skills/<name>/SKILL.md → skills, loaded on demand
  • hooks in settings → Claude-compatible hooks (exit 2 blocks)

Hooks, scripts and status lines from a repository wait for workspace trust. Plugins are validated, hash-pinned, re-verified on every start and capped at runtime. The full reference is docs/cli/extensibility.md.

👥 Multi-Agent Teams

Spawn parallel agents to divide and conquer large tasks.

# One quoted prompt per agent; each runs in its own git worktree
mc team spawn "Implement auth" "Implement payments" "Implement email" --max-agents 3
mc team status
mc team merge team-abc123

Command-line Command Families

These are the root registrations in src/main.ts (slash commands are a separate REPL interface). Run mc --help, mc <family> --help, and then mc <family> <subcommand> --help for exact arguments and nested flags.

| Root command/family | Purpose | |---------------------|---------| | mc auth | Login, logout, token, status and refresh | | mc status | Account and agent status | | mc doctor, mc setup | Health checks and setup wizard | | mc update, mc telemetry [action], mc bug [description] | Updates, telemetry and bug reports | | mc config | Layered settings, origins and migration | | mc index [path], mc context | Workspace indexing and semantic retrieval | | mc mcp | Server configuration, approval, OAuth, browsing and serving tools | | mc import <from>, mc plugin | Import other tools' setups and manage plugins | | mc agents, mc team, mc worktrees | Agent dashboard, parallel teams and worktree lifecycle | | mc triggers, mc schedule | Event-driven triggers and scheduled runs | | mc completions <shell> | Shell completion scripts | | mc sessions | List, inspect, export, fork, delete and prune saved conversations | | mc jobs | Background job status, logs, cancellation, retry, diff and pull | | mc cloud | Persistent remote workspaces (distinct from one-shot jobs) | | mc messages, mc join <sessionId>, mc share | Cross-session messages and live collaboration | | mc hooks, mc trust | Lifecycle hooks and execution trust records | | mc models, mc usage | Model catalog/pricing and account or session usage | | mc init, mc memory, mc rules, mc dream | Project conventions and memory consolidation | | mc review, mc bugfind [target] | Code review and bug scanning | | mc trace, mc benchmark | Execution traces and benchmarks | | mc marketplace, mc design | Skill marketplace and local design artboard | | mc ci (hidden integration family) | GitHub Actions protocol helpers; used by the action |

Root flags also include --json-schema <file>, --append-system-prompt <text>, --system-prompt-file <file>, --settings <fileOrJson>, --ignore-repo-settings, --mcp-config <files...>, --strict-mcp-config, --add-dir <dirs...> and the resume/fork and stdio permission flags described below. MCP config files passed explicitly are trusted for that run; --strict-mcp-config requires at least one such file. See mc --help and headless, permissions and extensibility for constraints. Nested commands have their own options: for example, mc sessions export --help and mc mcp serve --help describe different output/permission flags than the root.

Output Formats

| Format | Use Case | |--------|----------| | text | The final answer (default) | | json | One result object: answer, cost, usage, session id, denials | | stream-json | Every event, one per line (the protocol the SDKs speak) | | jsonl | Legacy event stream | | diff | Unified diff (git apply compatible) | | markdown | Clean markdown (for PR descriptions) | | github | GitHub Actions output format | | azure | Azure DevOps format | | sarif | SARIF 2.1.0 for code scanning | | stream | Raw event stream |

Permissions & Effort

What the agent may do (permission mode + rules) and how hard the model thinks (effort) are separate settings.

mc --permission-mode acceptEdits "refactor utils"     # default | acceptEdits | auto | plan | bypass
mc -p "fix tests" --headless --allowed-tools "run_command(npm test:*)" "edit_file(src/**)"
mc --effort high "find the race condition"            # low | medium | high

In a session the current mode is always shown at the bottom of the screen (and as [Manual] etc. on the prompt in the plain REPL):

| Mode | Stored as | What runs without asking | |------|-----------|--------------------------| | Manual | default | Reads only; every edit and command asks | | Accept edits | acceptEdits | File edits inside the workspace; commands still ask | | Auto | auto | Edits and commands. Still asks for destructive commands, anything outside the workspace, shell it can't verify and your ask rules | | Plan | plan | Nothing changes: research and propose a plan | | Bypass | bypass | Everything, destructive commands included; deny rules still apply. Only for throwaway environments |

/mode opens a picker; /mode auto, /mode accept-edits etc. switch directly, and --save keeps it as your starting mode. Shift+Tab cycles Manual → Accept edits → Auto → Plan (Bypass joins the cycle with cycleBypass, which mc setup offers). mc setup also picks the starting mode, or set permissions.defaultMode. Managed settings can disable Bypass (disableBypassPermissionsMode); Auto stays available. The old v1 settings key "mode": "auto" still means Manual. /plan jumps to plan mode, /effort high changes effort from the next request, and /permissions lists every effective rule and the file it came from. Print/headless runs are non-interactive: they do not open TTY approval dialogs. By default, calls that would ask are denied and reported in permissionDenials. The explicit exception is mc -p --permission-prompt stdio -o stream-json --input-format stream-json: a protocol client sends user messages on stdin and answers control_request permission messages with control_response messages (not terminal input). See the headless protocol for message shapes and timeouts. If only denials happened, the exit code is 6.

--auto-approve, bypass mode and background delegation are not blanket trust: explicit deny rules, protected/guarded operations and destructive-command checks still apply; these cannot be silently auto-approved. Prefer narrow allow rules for CI. Plugins and MCP tools remain subject to trust and permission checks, including when invoked by custom commands.

The rule syntax, evaluation order and shell-command handling are in docs/cli/permissions.md.

Configuration

mc config set model claude-sonnet                        # user settings
mc config set permissions.defaultMode acceptEdits --scope project
mc config list --show-origin                             # effective values + where each came from
mc config path                                           # settings file locations
mc config migrate                                        # fold legacy config.json/permissions.json/mcp.json
mc doctor                                                # auth, settings layers, invalid rules, defaults

Settings are layered. From highest to lowest precedence: managed policy → flags → env → .monkeyscode/settings.local.json (gitignored) → .monkeyscode/settings.json (committed) → ~/.monkeyscode/settings.json. Permission rules are combined across layers. For other keys, the highest-precedence layer wins. A repository can't set backend URLs or other trusted-only keys. Legacy config.json files are still read.

Shell Completions

mc completions bash >> ~/.bashrc
mc completions zsh >> ~/.zshrc
mc completions fish > ~/.config/fish/completions/mc.fish
mc completions pwsh >> $PROFILE

Slash Commands (REPL)

The 54 visible built-ins below follow the default registry and its full usageOf(c) strings. /help and the palette reflect the live registry.

Session

| Command | Usage | Description | |---------|-------|-------------| | /exit | /exit | Exit the REPL | | /clear | /clear | Start a fresh conversation; the old one stays resumable (Ctrl+L only clears the screen) | | /sessions | /sessions | List saved sessions | | /resume | /resume [id] | Resume a saved session | | /vim | /vim | Toggle Vim keybindings in the composer | | /logout | /logout | Sign out and exit | | /copy | /copy [n] | Copy the last reply (or the n-th last) | | /export | /export [file] | Save the conversation as Markdown | | /rewind | /rewind | Roll back conversation and files to an earlier prompt (Esc Esc) |

Model & Mode

| Command | Usage | Description | |---------|-------|-------------| | /model | /model [id] | Show or switch the model | | /mode | /mode [manual\|accept-edits\|auto\|plan\|bypass] [--save] | Show (picker) or switch the permission mode; --save makes it the starting mode | | /plan | /plan | Research and propose without edits | | /ask | /ask | Read-only Q&A about the codebase | | /effort | /effort [low\|medium\|high\|default] | Show or set reasoning effort | | /output-style | /output-style [name] [--save] | Show or switch output style |

Context & Stats

| Command | Usage | Description | |---------|-------|-------------| | /cost | /cost [all] | Session token usage and cost, per turn | | /context | /context [file] | Show context usage or add a file | | /compact | /compact ["focus"] | Compact the conversation to free context | | /status | /status | Model, permissions, effort, MCP and usage |

Git & Files

| Command | Usage | Description | |---------|-------|-------------| | /files | /files | List files read and modified this session | | /diff | /diff [session\|staged] [--plain] | Open Diff; --plain prints git diff | | /undo | /undo [--force] | Revert the last turn's file changes; keep conversation | | /redo | /redo [--force] | Re-apply changes reverted by undo | | /commit | /commit [message] | Stage everything and commit | | /review | /review [staged\|session\|<path>] [--chat] | Open Code Review or review in chat |

Agents & Delegation

| Command | Usage | Description | |---------|-------|-------------| | /agents | /agents [create <name> ["purpose"] \| edit <name> \| run <name> <task>] | List, create, edit or run custom subagents | | /goal | /goal "target" \| /goal status \| /goal cancel | Autonomous plan → execute → verify loop | | /loop | /loop [iterations] "command" | Iterate until a command passes | | /subtask | /subtask "task" \| /subtask status \| /subtask cancel <id> | Spawn or manage isolated background subtasks | | /fork | /fork "task" | Isolated background task; no conversation inherited | | /delegate | /delegate "task 1" "task 2" [...] | Parallel agents and merged results | | /search | /search "query" | Web search via the agent | | /docs | /docs <topic> | Look up documentation |

Skills, Hooks & Permissions

| Command | Usage | Description | |---------|-------|-------------| | /permissions | /permissions [show \| allow\|deny\|ask <rule> [--project] \| remove <rule> \| edit] | Effective rules and origins; add or remove rules | | /add-dir | /add-dir <path> [--save] | Allow another directory this session | | /skill | /skill [create <name>] | List or create skills | | /hooks | /hooks [create <name>] | List or create hooks | | /trust | /trust [revoke] | Review and trust workspace hooks/status line | | /setup | /setup [step] | Setup wizard | | /terminal-setup | /terminal-setup | Configure Shift+Enter and Option-as-Meta | | /shells | /shells [id \| kill <id\|all>] | List, inspect or stop background shells | | /mcp | /mcp [tools\|approve\|reconnect\|disable\|enable] [server] | Inspect/control session MCP servers | | /sandbox | /sandbox [on\|off\|status] [--save] | Turn OS sandboxing of shell commands on or off for this session; --save makes it the default for new sessions | | /plugins | /plugins [list \| browse [query] \| install <author>/<name>[@version] [--enable] \| info <name> \| enable <name> \| disable <name> \| update <name> \| rollback <name> \| remove <name> \| reload] | Find, install, review, enable, update and remove plugins without leaving the session. enable on a new plugin shows its capabilities and asks before trusting it; disable is session-only |

Project & Memory

| Command | Usage | Description | |---------|-------|-------------| | /init | /init | Initialize rules, permissions, skills and hooks | | /memory | /memory [show\|add [--user] "rule"\|edit [project\|user\|local]] | Show memory, add rules or edit a memory file | | /start | /start [number \| dismiss] | Suggested first tasks for this project: explain it, fix a bug in a file, write tests for a file. Alias: /starter |

Cloud & Messaging

| Command | Usage | Description | |---------|-------|-------------| | /cloud | /cloud | Hand current task off to a hosted cloud job | | /message | /message <session-id> "text" | Message another session | | /broadcast | /broadcast "text" | Message all active sessions |

Info & Help

| Command | Usage | Description | |---------|-------|-------------| | /bug | /bug | Diagnostic context for a bug report | | /doctor | /doctor | System health check | | /config | /config [key [value]] | Show/change user settings | | /notifications | /notifications [on\|off\|test\|status] | OS alerts when a task finishes or needs you (aliases /notify, /alerts) | | /update | /update [check] | Check for or install updates | | /help | /help | Show live help |

Custom (project, user, plugins, MCP)

| Command | Usage | Description | |---------|-------|-------------| | /commands | /commands [reload] | List and reload custom commands |

Aliases — each row maps every registered alias to its canonical command (name without the leading slash). Aliases accept the same arguments and are not additional built-ins.

| Canonical name | Aliases | |----------------|---------| | exit | /quit, /q | | clear | /cls, /new | | resume | /continue | | model | /models | | output-style | /style | | context | /ctx | | subtask | /sub | | delegate | /del | | permissions | /allowed-tools | | skill | /skills | | hooks | /hook | | shells | /bashes, /bg | | plugins | /plugin (legacy alias until the next release) | | bug | /feedback | | config | /settings | | update | /upgrade | | help | /h, /? |

Aliases (for example /h, /quit, /ctx, /models, /skills, /plugin) resolve to their primary commands, not additional built-ins. Project/user Markdown commands, enabled plugins and connected MCP prompts can add dynamic entries; /help shows their sources and name conflicts (built-ins win). Use /commands reload after editing command files and /plugins reload to refresh extensions. Custom command shell expansions and their tool scopes remain permission-gated; enabling a plugin does not bypass MCP trust or tool rules.

/plugins defaults to list; browse searches the configured public catalog (default app: cli; --app accepts ide, agent or cli). info <name> shows capabilities and version trust, and reload refreshes extensions. enable <name> and disable <name> change only the current session, never persistent settings or trust. Install and review version-pinned trust explicitly with mc plugin install and mc plugin enable; no slash command automatically grants trust. /plugin is the legacy Markdown command alias until the next release; migrate examples to /plugins.

Local authoring commands are mc plugin init [dir], mc plugin validate [dir] and mc plugin migrate-skills <from> <to> (copies skills without changing originals or granting trust). mc plugin remove <name> also accepts mc plugin uninstall <name>. Catalog discovery uses https://marketplace.monkeyscode.com unless MONKEYSCODE_CATALOG_ORIGIN overrides it; an empty value reports unavailability rather than sending a remote request. Plugin publishing remains unavailable until Phase 6.

Two different kinds of fork: /fork "task" starts an isolated background task without inheriting conversation history. mc sessions fork <id> copies a saved conversation to a new id. mc -r <id> --fork-session (or mc -c --fork-session) resumes history in a new session, leaving the original untouched; it is not a background task.

Type / to open the commands menu: it filters as you type and shows 5 rows at a time (↑↓ move, PgUp/PgDn jump, Tab completes, Enter runs the highlighted command, Esc closes). After a command and a space, the same menu lists its options: /model opu shows the matching models with their names and context size, and Enter switches to the highlighted one.

Tabs

| Key | Tab | |-----|-----| | ⌥1 / Alt+1 | Chat | | ⌥2 / Alt+2 · /diff | Diff: review, stage, discard, commit | | ⌥3 / Alt+3 · /review | Code Review | | Ctrl+] | Next tab (works in every terminal) |

The ⌥ keys work with your terminal's default settings (Terminal.app, iTerm2, VS Code, kitty, WezTerm, Ghostty, Windows Terminal). You can click a tab and scroll with the wheel in every tab: the wheel scrolls whichever pane is under the pointer (chat history, the Changes column, Diff, Code Review).

Selecting and copying: drag with the mouse to select text. The selection stays inside the pane where you started (the chat history, the Changes column or the input box), so selecting the chat never picks up the diff beside it. Hold the drag above or below the chat or the Changes column to scroll it and keep selecting (further from the edge scrolls faster). Selected text keeps its colours, drawn in inverse. It's copied to the clipboard when you release the button, and a notice confirms it. Any key clears the highlight. Hold Shift (⌥ in iTerm2 and Terminal.app) while dragging for the terminal's own selection, which spans the whole screen width. /copy copies the last answer. Set mouse to tabs to keep the mouse to Diff / Code Review (the terminal selects in Chat, but the wheel no longer scrolls it), or to off to turn the mouse off.

In-app chat scrolling: The chat viewport is internally scrollable — the composer and status line stay fixed while you navigate history. Use the mouse wheel, PageUp/PageDown, or Ctrl+G to jump to the latest. Scrolling anchors survive streaming, tab switches, and terminal resizes. A "new output" indicator appears when you're scrolled up during a run.

Changes column: on terminals 120 columns or wider, the chat shows the session's changed files and their diffs on the right. Ctrl+F focuses it (←/→ file, ↑/↓ or the wheel to scroll), Ctrl+Y hides or shows it, and changesPane: false starts without it. /diff opens the full Diff tab.

Multi-line input: Press Shift+Enter for a newline (or Alt+Enter, or end a line with \). Enter sends the message. Run mc setup to configure your terminal for distinguishable Shift+Enter if it isn't already.

SDK

Run this agent from your own code, with your settings, rules, hooks and MCP servers:

import { query } from '@monkeyscode/agent';   // the engine in process (Node)

const q = query({ prompt: 'Fix the failing test', options: { allowedTools: ['run_command(npm test:*)'], maxCostUsd: 1 } });
for await (const e of q) if (e.type === 'result') console.log(e.result);

From Python (monkeyscode.cli_process), Go (monkeyscode.Query) or TypeScript (@monkeyscode/sdk/subprocess), the same API drives an installed mc over stdio. Guide: docs/cli/sdk.md.

The cloud clients (MonkeysCode in @monkeyscode/sdk) call the hosted API instead.

Updates

Version checks are quiet by default — normal TUI and plain-REPL sessions never phone the registry or announce available updates. Use the explicit command when you want to check or upgrade:

mc update --check     # Check for a newer release (exits nonzero on failure)
mc update              # Upgrade via npm, Homebrew, winget, Chocolatey, Scoop, or binary download
mc update --version 1.2.8   # Install a specific version (npm or binary)
mc update --rollback   # Revert to the previous binary (standalone installs only)

Binary updates verify a SHA-256 checksum before replacement and keep a .backup for rollback. On Windows, the replacement is staged and completes when you exit the CLI (the running process can't overwrite its own executable).

Accessibility

Screen reader-friendly mode (no ANSI codes, no animations):

mc --accessible "fix the bug"
# or
export MONKEYSCODE_ACCESSIBLE=1

Telemetry

Telemetry is enabled by default and collects anonymous usage data to improve the product.

mc telemetry status    # Check status
mc telemetry disable   # Opt out
mc telemetry show      # See what's collected

Respects the DO_NOT_TRACK=1 environment variable.

Documentation

License

MIT © MonkeysCloud Corp