Enum Commands 

pub enum Commands {
    Run {
        task: String,
        json: bool,
    },
    Plan {
        task: String,
        json: bool,
    },
    Review {
        base: Option<String>,
        paths: Vec<String>,
        dry_run: bool,
    },
    Chat,
    Tui,
    Launch {
        port: u16,
        tui: bool,
        pro: bool,
    },
    Console {
        port: u16,
    },
    Fix {
        problem: Vec<String>,
    },
    Explique {
        target: Vec<String>,
    },
    Annule {
        id: Option<String>,
        tout: bool,
    },
    Bonjour,
    Budget {
        amount: Option<String>,
    },
    Idees {
        filter: Vec<String>,
    },
    Whatis {
        term: Vec<String>,
    },
    Mode {
        mode: Option<String>,
    },
    Daemon,
    Agent {
        action: AgentAction,
    },
    Swarm {
        task: String,
    },
    Schedule {
        task: String,
        cron: String,
        autonomy: Option<String>,
        report: Vec<String>,
    },
    Model {
        set: Option<String>,
        list: bool,
    },
    Route {
        action: RouteAction,
    },
    Auth {
        action: AuthAction,
    },
    Skills {
        action: SkillsAction,
    },
    Plugins {
        action: PluginsAction,
    },
    Tools {
        action: ToolsAction,
    },
    Security {
        action: SecurityAction,
    },
    Github {
        action: GithubAction,
    },
    Compact {
        task: Option<String>,
        out: Option<PathBuf>,
        json: bool,
    },
    Mcp {
        action: McpAction,
    },
    Checkpoint {
        action: CheckpointAction,
    },
    Rewind {
        id: String,
    },
    Replay {
        run_id: String,
        scrub: bool,
    },
    Gateway {
        action: GatewayAction,
    },
    Sessions {
        action: SessionAction,
    },
    Learn,
    Init,
    Status,
    Memory {
        action: MemoryAction,
    },
    Permissions {
        action: PermissionAction,
    },
    Profile {
        action: ProfileAction,
    },
    Import {
        source: ImportSource,
    },
    Config {
        edit: bool,
    },
    Update,
    Doctor,
    Setup,
    Demo,
    Share,
    Hook {
        action: HookAction,
    },
    Voice {
        action: VoiceAction,
    },
    Browser {
        url: String,
    },
}

Variants

Run

Run a single agentic task

Fields

task: String

Task description

json: bool

Emit NDJSON event stream (same as the global –json flag, but may follow the task: sparrow run "..." --json)

Plan

Create a read-only execution plan for a task

Fields

task: String

Task description

json: bool

Emit JSON instead of Markdown

Review

Adversarial review of the current local diff (uncommitted, staged, and commits ahead of --base). Read-only — no edits, no commits, no network beyond the model call. Findings are structured around security, correctness, regressions, performance, and readability.

Fields

base: Option<String>

Base ref to diff against (defaults to origin/main, then main, then HEAD~1).

paths: Vec<String>

Only review changes touching these path globs (repeatable).

dry_run: bool

Print the prompt the model will see and exit (no model call).

Chat

Interactive multi-turn chat

Tui

Launch TUI

Launch

Launch first-run setup, then the WebView cockpit

Fields

port: u16

TCP port for the WebView HTTP/WS server

tui: bool

Launch the terminal TUI instead of the WebView cockpit

pro: bool

Use the older expert setup wizard before opening the surface

Console

Launch webview console (HTTP + WebSocket)

Fields

port: u16

TCP port for the webview HTTP/WS server

Fix

Réparer un problème — décris ce qui ne va pas, Sparrow diagnostique puis corrige (avec ton accord). « sparrow fix “message d’erreur” », ou sans argument pour scanner le dossier courant.

Fields

problem: Vec<String>

Le problème, avec tes mots, ou une erreur collée (entre guillemets si elle contient des espaces). Optionnel : sans argument, Sparrow inspecte le dossier courant.

Explique

Expliquer un fichier, une erreur ou un concept en langage simple. « sparrow explique src/main.rs » · « sparrow explique “borrow checker” »

Fields

target: Vec<String>

Ce qu’il faut expliquer : un chemin de fichier, une erreur, ou un mot (entre guillemets si plusieurs mots).

Annule

Annuler la dernière action de Sparrow — revient au dernier point de sauvegarde, rien n’est perdu. « sparrow annule » · « sparrow annule –tout » pour revenir au début de la session.

Fields

id: Option<String>

Point de sauvegarde précis (sinon : le tout dernier).

tout: bool

Revenir au tout premier point de sauvegarde de la session.

Bonjour

Dire bonjour — l’accueil chaleureux : Sparrow regarde ton dossier et te propose quoi faire. Parfait pour un premier contact.

Budget

Voir ou changer le plafond de dépense par session. « sparrow budget » affiche le réglage actuel ; « sparrow budget 2€ » le change.

Fields

amount: Option<String>

Le montant max par session (ex. « 2€ », « $0.50 », « 1.5 »). Vide : affiche le réglage actuel.

Idees

Des idées de ce que tu peux faire avec Sparrow, classées par profil. « sparrow idees » · « sparrow idees enseignant » · « sparrow idees “factures” ».

Fields

filter: Vec<String>

Filtre : un profil (enseignant, developpeur, …) ou un mot-clé.

Whatis

C’est quoi ce mot ? — définition instantanée d’un terme de Sparrow, en deux phrases simples, sans appel modèle. « sparrow whatis token ».

Fields

term: Vec<String>

Le terme à définir (ex. checkpoint, token, swarm). Vide : liste les mots connus.

Mode

Choisir comment Sparrow te parle : simple (langage clair, zéro jargon), pro (sortie technique complète) ou auto. Sans argument : affiche le mode actuel.

Fields

mode: Option<String>

« simple », « pro » ou « auto ».

Daemon

Run headless Sparrow runtime daemon

Agent

Manage persistent agents

Fields

action: AgentAction

Swarm

Run swarm: planner → coder → verifier

Fields

task: String

Task or plan file

Schedule

Schedule periodic jobs

Fields

task: String

Task description

cron: String

Cron expression

autonomy: Option<String>

Autonomy level for scheduled jobs

report: Vec<String>

Report to surfaces

Model

Manage model routing

Fields

set: Option<String>

Set active route

list: bool

List available models

Route

Configure intelligent auto-routing provider

Fields

action: RouteAction

Auth

Manage provider credentials

Fields

action: AuthAction

Skills

Manage skill library

Fields

action: SkillsAction

Plugins

Manage local Sparrow plugins

Fields

action: PluginsAction

Tools

Inspect and gate toolsets

Fields

action: ToolsAction

Security

Security audit of config, permissions, plugins, hooks, secrets

Fields

action: SecurityAction

Github

GitHub Action / remote PR workflow

Fields

action: GithubAction

Compact

Compact context and write a durable handoff doc

Fields

task: Option<String>

Task description (recorded in the handoff)

out: Option<PathBuf>

Output path (default: .sparrow/handoff/.md)

json: bool

Emit JSON instead of Markdown to stdout (the file is always Markdown)

Mcp

Manage MCP connectors

Fields

action: McpAction

Checkpoint

List checkpoints

Fields

action: CheckpointAction

Rewind

Rewind to a checkpoint

Fields

id: String

Checkpoint ID or number

Replay

Replay a transcript

Fields

run_id: String

Run ID to replay

scrub: bool

Open an interactive TUI scrubber (←/→ to step through events)

Gateway

Start/stop gateway daemon

Fields

action: GatewayAction

Sessions

Manage saved sessions

Fields

action: SessionAction

Learn

Interactive tutorial

Init

Initialize a project with .sparrow/ config

Status

Show live status (active runs, budget, session)

Memory

Manage persistent memory

Fields

action: MemoryAction

Permissions

Inspect and update permission policy

Fields

action: PermissionAction

Profile

Profile management

Fields

action: ProfileAction

Import

Import config from another tool (claude-code, codex, opencode, openclaw)

Fields

source: ImportSource

Config

Edit configuration

Fields

edit: bool

Open config.toml in editor

Update

Self-update

Doctor

Run diagnostics

Setup

(Re)run conversational setup

Demo

Run a self-contained demo (snake game)

Share

Share latest session as GitHub Gist

Hook

Install or scan security pre-commit hooks

Fields

action: HookAction

Voice

Voice commands (speak, transcribe, providers)

Fields

action: VoiceAction

Browser

Test browser/vision (screenshot, navigate)

Fields

url: String

URL to test

Trait Implementations

impl FromArgMatches for Commands

fn from_arg_matches(__clap_arg_matches: &ArgMatches) -> Result<Self, Error>

Instantiate Self from ArgMatches, parsing the arguments as needed.

fn from_arg_matches_mut( __clap_arg_matches: &mut ArgMatches, ) -> Result<Self, Error>

Instantiate Self from ArgMatches, parsing the arguments as needed.

fn update_from_arg_matches( &mut self, __clap_arg_matches: &ArgMatches, ) -> Result<(), Error>

Assign values from ArgMatches to self.

fn update_from_arg_matches_mut<'b>( &mut self, __clap_arg_matches: &mut ArgMatches, ) -> Result<(), Error>

Assign values from ArgMatches to self.

impl Subcommand for Commands

fn augment_subcommands<'b>(__clap_app: Command) -> Command

Append to Command so it can instantiate Self via FromArgMatches::from_arg_matches_mut

fn augment_subcommands_for_update<'b>(__clap_app: Command) -> Command

Append to Command so it can instantiate self via FromArgMatches::update_from_arg_matches_mut

fn has_subcommand(__clap_name: &str) -> bool

Test whether Self can parse a specific subcommand