eliferjunior/fig

Amazon Q (formerly Fig) — Terminal Autocomplete & CLI Tools

Overview

Amazon Q Developer (formerly Fig), the terminal tool that provides IDE-style autocomplete, AI chat, and CLI builder capabilities. Helps developers create custom completion specs, build CLI tools with autocomplete, and configure terminal productivity features.

Instructions

Custom Completion Specs

Define autocomplete for your own CLI tools:

// src/my-cli.ts — Completion spec for a custom CLI tool
const completionSpec: Fig.Spec = {
  name: "deploy",
  description: "Deploy services to cloud infrastructure",
  subcommands: [
    {
      name: "create",
      description: "Create a new deployment",
      args: {
        name: "service",
        description: "Service name to deploy",
        // Dynamic suggestions from an API or command output
        generators: {
          script: ["bash", "-c", "ls services/"],
          postProcess: (output) =>
            output.split("\n").filter(Boolean).map((name) => ({
              name,
              description: `Deploy ${name} service`,
              icon: "📦",
            })),
        },
      },
      options: [
        {
          name: ["--env", "-e"],
          description: "Target environment",
          args: {
            name: "environment",
            suggestions: [
              { name: "staging", description: "Staging environment", icon: "🟡" },
              { name: "production", description: "Production environment", icon: "🔴" },
              { name: "development", description: "Development environment", icon: "🟢" },
            ],
          },
          isRequired: true,
        },
        {
          name: ["--replicas", "-r"],
          description: "Number of replicas",
          args: { name: "count", default: "2" },
        },
        {
          name: "--dry-run",
          description: "Preview changes without deploying",
        },
        {
          name: ["--tag", "-t"],
          description: "Docker image tag",
          args: {
            name: "tag",
            generators: {
              // Suggest recent git tags
              script: ["bash", "-c", "git tag --sort=-version:refname | head -10"],
              postProcess: (output) =>
                output.split("\n").filter(Boolean).map((tag) => ({
                  name: tag,
                  icon: "🏷️",
                })),
            },
          },
        },
      ],
    },
    {
      name: "status",
      description: "Check deployment status",
      args: {
        name: "service",
        isOptional: true,
        generators: {
          script: ["bash", "-c", "kubectl get deployments -o jsonpath='{.items[*].metadata.name}'"],
          splitOn: " ",
        },
      },
      options: [
        {
          name: ["--watch", "-w"],
          description: "Watch status in real-time",
        },
        {
          name: "--format",
          description: "Output format",
          args: {
            suggestions: ["table", "json", "yaml"],
          },
        },
      ],
    },
    {
      name: "rollback",
      description: "Rollback to previous version",
      isDangerous: true,    // Shows warning icon
      args: {
        name: "service",
        generators: {
          script: ["bash", "-c", "kubectl get deployments -o jsonpath='{.items[*].metadata.name}'"],
          splitOn: " ",
        },
      },
      options: [
        {
          name: "--revision",
          description: "Specific revision to rollback to",
          args: {
            name: "revision",
            generators: {
              // Dynamic: list revision history for the selected service
              script: ({ tokens }) => {
                const service = tokens[2]; // deploy rollback <service>
                return ["bash", "-c", `kubectl rollout history deployment/${service} | tail -n +3 | awk '{print $1}'`];
              },
              postProcess: (output) =>
                output.split("\n").filter(Boolean).map((rev) => ({
                  name: rev,
                  description: `Revision ${rev}`,
                })),
            },
          },
        },
      ],
    },
    {
      name: "logs",
      description: "View deployment logs",
      args: {
        name: "service",
        generators: {
          script: ["bash", "-c", "kubectl get deployments -o jsonpath='{.items[*].metadata.name}'"],
          splitOn: " ",
        },
      },
      options: [
        { name: ["--follow", "-f"], description: "Stream logs in real-time" },
        { name: "--since", description: "Show logs since duration", args: { suggestions: ["1m", "5m", "1h", "24h"] } },
        { name: "--tail", description: "Number of recent lines", args: { name: "lines", default: "100" } },
      ],
    },
  ],
  options: [
    {
      name: ["--verbose", "-v"],
      description: "Enable verbose output",
      isPersistent: true,    // Available on all subcommands
    },
    {
      name: "--config",
      description: "Path to config file",
      isPersistent: true,
      args: { template: "filepaths" },   // File path autocomplete
    },
  ],
};

export default completionSpec;

Dotfile Scripts

Create portable shell scripts that run across machines:

# ~/.fig/scripts/setup-project.sh — Reusable project setup script
#!/bin/bash
# Fig script: Initialize a new TypeScript project with standard tooling

set -euo pipefail

PROJECT_NAME=${1:?"Usage: fig run setup-project <name>"}

echo "📁 Creating project: $PROJECT_NAME"
mkdir -p "$PROJECT_NAME" && cd "$PROJECT_NAME"

# Initialize with package.json
cat > package.json << EOF
{
  "name": "$PROJECT_NAME",
  "version": "0.1.0",
  "type": "module",
  "scripts": {
    "dev": "tsx watch src/index.ts",
    "build": "tsup src/index.ts",
    "lint": "biome check .",
    "format": "biome format --write .",
    "test": "vitest"
  }
}
EOF

# Install dependencies
pnpm install typescript tsx tsup vitest @biomejs/biome

# TypeScript config
cat > tsconfig.json << 'EOF'
{
  "compilerOptions": {
    "target": "ES2022",
    "module": "ESNext",
    "moduleResolution": "bundler",
    "strict": true,
    "outDir": "dist",
    "rootDir": "src",
    "declaration": true
  },
  "include": ["src"]
}
EOF

# Biome config
cat > biome.json << 'EOF'
{
  "$schema": "https://biomejs.dev/schemas/1.9.4/schema.json",
  "organizeImports": { "enabled": true },
  "linter": { "enabled": true },
  "formatter": { "indentStyle": "space", "indentWidth": 2 }
}
EOF

# Create entry point
mkdir -p src
cat > src/index.ts << 'EOF'
console.log("Hello from $PROJECT_NAME!");
EOF

# Git init
git init
echo "node_modules/\ndist/\n.env" > .gitignore

echo "✅ Project $PROJECT_NAME created!"
echo "   cd $PROJECT_NAME && pnpm dev"

SSH Integration

Auto-complete for remote servers:

// src/ssh-hosts.ts — Dynamic SSH host suggestions
const sshSpec: Fig.Spec = {
  name: "ssh",
  args: {
    name: "destination",
    generators: {
      // Parse SSH config for host suggestions
      script: ["bash", "-c", `
        grep "^Host " ~/.ssh/config 2>/dev/null | awk '{print $2}' | grep -v '*'
      `],
      postProcess: (output) =>
        output.split("\n").filter(Boolean).map((host) => ({
          name: host,
          description: "SSH host from config",
          icon: "🖥️",
          priority: 80,
        })),
    },
  },
};

Installation

# macOS
brew install amazon-q

# Or direct download from https://aws.amazon.com/q/developer/

# Login and configure
q login
q settings

Examples

Example 1: Setting up Fig with a custom configuration

User request:

I just installed Fig. Help me configure it for my TypeScript + React workflow with my preferred keybindings.

The agent creates the configuration file with TypeScript-aware settings, configures relevant plugins/extensions for React development, sets up keyboard shortcuts matching the user's preferences, and verifies the setup works correctly.

Example 2: Extending Fig with custom functionality

User request:

I want to add a custom dotfile scripts to Fig. How do I build one?

The agent scaffolds the extension/plugin project, implements the core functionality following Fig's API patterns, adds configuration options, and provides testing instructions to verify it works end-to-end.

Guidelines

1. Generators for dynamic suggestions — Use shell scripts to generate suggestions from live data (git branches, k8s resources, file lists)
2. Mark dangerous commands — Set isDangerous: true on destructive subcommands (delete, drop, rollback)
3. Use templates for common arg types — template: "filepaths" for file args, template: "folders" for directories
4. Persistent options — Use isPersistent: true for flags like --verbose that apply to all subcommands
5. Icons and descriptions — Add icons and descriptions to suggestions; visual cues speed up selection
6. Publish to the spec repo — Share your completion specs via the fig/autocomplete GitHub repo for community use
7. Test with fig settings developer.enabled true — Enable developer mode for spec debugging and hot reload
8. Dotfile scripts for portability — Store setup scripts in ~/.fig/scripts/; they sync across machines via Fig's dotfile management