<!DOCTYPE html> <html lang="en"> <head> <link rel="stylesheet" href="https://unpkg.com/latex.css/style.min.css" /> <link rel="stylesheet" href="/assets/main.css" /> <meta charset="utf-8"> <meta name="viewport" content="width=device-width, initial-scale=1.0"> <title>Interacting with Siri using the command line</title> <meta name="og:site_name" content="Navan Chauhan" /> <link rel="canonical" href="https://web.navan.dev/posts/2023-02-08-Interact-with-siri-from-the-terminal.html" /> <meta name="twitter:url" content="https://web.navan.dev/posts/2023-02-08-Interact-with-siri-from-the-terminal.html /> <meta name="og:url" content="https://web.navan.dev/posts/2023-02-08-Interact-with-siri-from-the-terminal.html" /> <meta name="twitter:title" content="Interacting with Siri using the command line" /> <meta name="og:title" content="Interacting with Siri using the command line" /> <meta name="description" content="Code snippet to interact with Siri by issuing commands from the command-line." /> <meta name="twitter:description" content="Code snippet to interact with Siri by issuing commands from the command-line." /> <meta name="og:description" content="Code snippet to interact with Siri by issuing commands from the command-line." /> <meta name="twitter:card" content="summary_large_image" /> <meta name="viewport" content="width=device-width, initial-scale=1.0" /> <link rel="shortcut icon" href="/images/favicon.png" type="image/png" /> <link rel="alternate" href="/feed.rss" type="application/rss+xml" title="Subscribe to Navan Chauhan" /> <meta name="twitter:image" content="https://web.navan.dev/images/opengraph/posts/2023-02-08-Interact-with-siri-from-the-terminal.png" /> <meta name="og:image" content="https://web.navan.dev/images/opengraph/posts/2023-02-08-Interact-with-siri-from-the-terminal.png" /> <meta name="google-site-verification" content="LVeSZxz-QskhbEjHxOi7-BM5dDxTg53x2TwrjFxfL0k" /> <script data-goatcounter="https://navanchauhan.goatcounter.com/count" async src="//gc.zgo.at/count.js"></script> <script defer data-domain="web.navan.dev" src="https://plausible.io/js/plausible.js"></script> <link rel="manifest" href="/manifest.json" /> </head> <body> <center><nav style="display: block;"> | <a href="/">home</a> | <a href="/about/">about/links</a> | <a href="/posts/">posts</a> | <a href="/3D-Designs/">3D designs</a> | <!--<a href="/publications/">publications</a> |--> <!--<a href="/repo/">iOS repo</a> |--> <a href="/feed.rss">RSS Feed</a> | </nav> </center> <main> <h1>Interacting with Siri using the command line</h1> <p>My main objective was to see if I could issue multi-intent commands in one go. Obviously, Siri cannot do that (neither can Alexa, Cortana, or Google Assistant). The script here can issue either a single command, or use the help of OpenAI's DaVinci model to extract multiple commands and pass them onto siri.</p> <h2>Prerequisites</h2> <ul> <li>Run macOS</li> <li>Enable type to Siri (Settings > Accessibility -> Type to Siri)</li> <li>Enable the Terminal to control System Events (The first time you run the script, it will prompt you to enable it)</li> </ul> <h2>Show me ze code</h2> <p>If you are here just for the code:</p> <div class="codehilite"> <pre><span></span><code><span class="kn">import</span> <span class="nn">argparse</span> <span class="kn">import</span> <span class="nn">applescript</span> <span class="kn">import</span> <span class="nn">openai</span> <span class="kn">from</span> <span class="nn">os</span> <span class="kn">import</span> <span class="n">getenv</span> <span class="n">openai</span><span class="o">.</span><span class="n">api_key</span> <span class="o">=</span> <span class="n">getenv</span><span class="p">(</span><span class="s2">"OPENAI_KEY"</span><span class="p">)</span> <span class="n">engine</span> <span class="o">=</span> <span class="s2">"text-davinci-003"</span> <span class="k">def</span> <span class="nf">execute_with_llm</span><span class="p">(</span><span class="n">command_text</span><span class="p">:</span> <span class="nb">str</span><span class="p">)</span> <span class="o">-></span> <span class="kc">None</span><span class="p">:</span> <span class="n">llm_prompt</span> <span class="o">=</span> <span class="sa">f</span><span class="s2">"""You are provided with multiple commands as a single command. Break down all the commands and return them in a list of strings. If you are provided with a single command, return a list with a single string, trying your best to understand the command.</span> <span class="s2"> Example:</span> <span class="s2"> Q: "Turn on the lights and turn off the lights"</span> <span class="s2"> A: ["Turn on the lights", "Turn off the lights"]</span> <span class="s2"> Q: "Switch off the lights and then play some music"</span> <span class="s2"> A: ["Switch off the lights", "Play some music"]</span> <span class="s2"> Q: "I am feeling sad today, play some music"</span> <span class="s2"> A: ["Play some cheerful music"]</span> <span class="s2"> Q: "</span><span class="si">{</span><span class="n">command_text</span><span class="si">}</span><span class="s2">"</span> <span class="s2"> A: </span> <span class="s2"> """</span> <span class="n">completion</span> <span class="o">=</span> <span class="n">openai</span><span class="o">.</span><span class="n">Completion</span><span class="o">.</span><span class="n">create</span><span class="p">(</span><span class="n">engine</span><span class="o">=</span><span class="n">engine</span><span class="p">,</span> <span class="n">prompt</span><span class="o">=</span><span class="n">llm_prompt</span><span class="p">,</span> <span class="n">max_tokens</span><span class="o">=</span><span class="nb">len</span><span class="p">(</span><span class="n">command_text</span><span class="o">.</span><span class="n">split</span><span class="p">(</span><span class="s2">" "</span><span class="p">))</span><span class="o">*</span><span class="mi">2</span><span class="p">)</span> <span class="k">for</span> <span class="n">task</span> <span class="ow">in</span> <span class="nb">eval</span><span class="p">(</span><span class="n">completion</span><span class="o">.</span><span class="n">choices</span><span class="p">[</span><span class="mi">0</span><span class="p">]</span><span class="o">.</span><span class="n">text</span><span class="p">):</span> <span class="n">execute_command</span><span class="p">(</span><span class="n">task</span><span class="p">)</span> <span class="k">def</span> <span class="nf">execute_command</span><span class="p">(</span><span class="n">command_text</span><span class="p">:</span> <span class="nb">str</span><span class="p">)</span> <span class="o">-></span> <span class="kc">None</span><span class="p">:</span> <span class="w"> </span><span class="sd">"""Execute a Siri command."""</span> <span class="n">script</span> <span class="o">=</span> <span class="n">applescript</span><span class="o">.</span><span class="n">AppleScript</span><span class="p">(</span><span class="sa">f</span><span class="s2">"""</span> <span class="s2"> tell application "System Events" to tell the front menu bar of process "SystemUIServer"</span> <span class="s2"> tell (first menu bar item whose description is "Siri")</span> <span class="s2"> perform action "AXPress"</span> <span class="s2"> end tell</span> <span class="s2"> end tell</span> <span class="s2"> delay 2</span> <span class="s2"> tell application "System Events"</span> <span class="s2"> set textToType to "</span><span class="si">{</span><span class="n">command_text</span><span class="si">}</span><span class="s2">"</span> <span class="s2"> keystroke textToType</span> <span class="s2"> key code 36</span> <span class="s2"> end tell</span> <span class="s2"> """</span><span class="p">)</span> <span class="n">script</span><span class="o">.</span><span class="n">run</span><span class="p">()</span> <span class="k">if</span> <span class="vm">__name__</span> <span class="o">==</span> <span class="s2">"__main__"</span><span class="p">:</span> <span class="n">parser</span> <span class="o">=</span> <span class="n">argparse</span><span class="o">.</span><span class="n">ArgumentParser</span><span class="p">()</span> <span class="n">parser</span><span class="o">.</span><span class="n">add_argument</span><span class="p">(</span><span class="s2">"command"</span><span class="p">,</span> <span class="n">nargs</span><span class="o">=</span><span class="s2">"?"</span><span class="p">,</span> <span class="nb">type</span><span class="o">=</span><span class="nb">str</span><span class="p">,</span> <span class="n">help</span><span class="o">=</span><span class="s2">"The command to pass to Siri"</span><span class="p">,</span> <span class="n">default</span><span class="o">=</span><span class="s2">"What time is it?"</span><span class="p">)</span> <span class="n">parser</span><span class="o">.</span><span class="n">add_argument</span><span class="p">(</span><span class="s1">'--openai'</span><span class="p">,</span> <span class="n">action</span><span class="o">=</span><span class="n">argparse</span><span class="o">.</span><span class="n">BooleanOptionalAction</span><span class="p">,</span> <span class="n">help</span><span class="o">=</span><span class="s2">"Use OpenAI to detect multiple intents"</span><span class="p">,</span> <span class="n">default</span><span class="o">=</span><span class="kc">False</span><span class="p">)</span> <span class="n">args</span> <span class="o">=</span> <span class="n">parser</span><span class="o">.</span><span class="n">parse_args</span><span class="p">()</span> <span class="k">if</span> <span class="n">args</span><span class="o">.</span><span class="n">openai</span><span class="p">:</span> <span class="n">execute_with_llm</span><span class="p">(</span><span class="n">args</span><span class="o">.</span><span class="n">command</span><span class="p">)</span> <span class="k">else</span><span class="p">:</span> <span class="n">execute_command</span><span class="p">(</span><span class="n">args</span><span class="o">.</span><span class="n">command</span><span class="p">)</span> </code></pre> </div> <p>Usage:</p> <div class="codehilite"> <pre><span></span><code>python3<span class="w"> </span>main.py<span class="w"> </span><span class="s2">"play some taylor swift"</span> python3<span class="w"> </span>main.py<span class="w"> </span><span class="s2">"turn off the lights and play some music"</span><span class="w"> </span>--openai </code></pre> </div> <h2>ELI5</h2> <p>I am not actually going to explain it as if I am explaining to a five-year old kid.</p> <h3>AppleScript</h3> <p>In the age of Siri Shortcuts, AppleScript can still do more. It is a scripting language created by Apple that can help you automate pretty much anything you see on your screen.</p> <p>We use the following AppleScript to trigger Siri and then type in our command:</p> <div class="codehilite"> <pre><span></span><code><span class="k">tell</span> <span class="nb">application</span> <span class="s2">"System Events"</span> <span class="k">to</span> <span class="k">tell</span> <span class="nb">the</span> <span class="nb">front</span> <span class="na">menu</span> <span class="nv">bar</span> <span class="k">of</span> <span class="nv">process</span> <span class="s2">"SystemUIServer"</span> <span class="k">tell</span> <span class="p">(</span><span class="nb">first</span> <span class="na">menu</span> <span class="nv">bar</span> <span class="nb">item</span> <span class="nb">whose</span> <span class="nv">description</span> <span class="ow">is</span> <span class="s2">"Siri"</span><span class="p">)</span> <span class="nb">perform action</span> <span class="s2">"AXPress"</span> <span class="k">end</span> <span class="k">tell</span> <span class="k">end</span> <span class="k">tell</span> <span class="nb">delay</span> <span class="mi">2</span> <span class="k">tell</span> <span class="nb">application</span> <span class="s2">"System Events"</span> <span class="k">set</span> <span class="nv">textToType</span> <span class="k">to</span> <span class="s2">"Play some rock music"</span> <span class="nv">keystroke</span> <span class="nv">textToType</span> <span class="na">key code</span> <span class="mi">36</span> <span class="k">end</span> <span class="k">tell</span> </code></pre> </div> <p>This first triggers Siri, waits for a couple of seconds, and then types in our command. In the script, this functionality is handled by the <code>execute_command</code> function.</p> <div class="codehilite"> <pre><span></span><code><span class="kn">import</span> <span class="nn">applescript</span> <span class="k">def</span> <span class="nf">execute_command</span><span class="p">(</span><span class="n">command_text</span><span class="p">:</span> <span class="nb">str</span><span class="p">)</span> <span class="o">-></span> <span class="kc">None</span><span class="p">:</span> <span class="w"> </span><span class="sd">"""Execute a Siri command."""</span> <span class="n">script</span> <span class="o">=</span> <span class="n">applescript</span><span class="o">.</span><span class="n">AppleScript</span><span class="p">(</span><span class="sa">f</span><span class="s2">"""</span> <span class="s2"> tell application "System Events" to tell the front menu bar of process "SystemUIServer"</span> <span class="s2"> tell (first menu bar item whose description is "Siri")</span> <span class="s2"> perform action "AXPress"</span> <span class="s2"> end tell</span> <span class="s2"> end tell</span> <span class="s2"> delay 2</span> <span class="s2"> tell application "System Events"</span> <span class="s2"> set textToType to "</span><span class="si">{</span><span class="n">command_text</span><span class="si">}</span><span class="s2">"</span> <span class="s2"> keystroke textToType</span> <span class="s2"> key code 36</span> <span class="s2"> end tell</span> <span class="s2"> """</span><span class="p">)</span> <span class="n">script</span><span class="o">.</span><span class="n">run</span><span class="p">()</span> </code></pre> </div> <h3>Multi-Intent Commands</h3> <p>We can call OpenAI's API to autocomplete our prompt and extract multiple commands. We don't need to use OpenAI's API, and can also simply use Google's Flan-T5 model using HuggingFace's transformers library. </p> <h4>Ze Prompt</h4> <div class="codehilite"> <pre><span></span><code>You are provided with multiple commands as a single command. Break down all the commands and return them in a list of strings. If you are provided with a single command, return a list with a single string, trying your best to understand the command. Example: Q: "Turn on the lights and turn off the lights" A: ["Turn on the lights", "Turn off the lights"] Q: "Switch off the lights and then play some music" A: ["Switch off the lights", "Play some music"] Q: "I am feeling sad today, play some music" A: ["Play some cheerful music"] Q: "{command_text}" A: </code></pre> </div> <p>This prompt gives the model a few examples to increase the generation accuracy, along with instructing it to return a Python list. </p> <h4>Ze Code</h4> <div class="codehilite"> <pre><span></span><code><span class="kn">import</span> <span class="nn">openai</span> <span class="kn">from</span> <span class="nn">os</span> <span class="kn">import</span> <span class="n">getenv</span> <span class="n">openai</span><span class="o">.</span><span class="n">api_key</span> <span class="o">=</span> <span class="n">getenv</span><span class="p">(</span><span class="s2">"OPENAI_KEY"</span><span class="p">)</span> <span class="n">engine</span> <span class="o">=</span> <span class="s2">"text-davinci-003"</span> <span class="k">def</span> <span class="nf">execute_with_llm</span><span class="p">(</span><span class="n">command_text</span><span class="p">:</span> <span class="nb">str</span><span class="p">)</span> <span class="o">-></span> <span class="kc">None</span><span class="p">:</span> <span class="n">llm_prompt</span> <span class="o">=</span> <span class="sa">f</span><span class="s2">"""You are provided with multiple commands as a single command. Break down all the commands and return them in a list of strings. If you are provided with a single command, return a list with a single string, trying your best to understand the command.</span> <span class="s2"> Example:</span> <span class="s2"> Q: "Turn on the lights and turn off the lights"</span> <span class="s2"> A: ["Turn on the lights", "Turn off the lights"]</span> <span class="s2"> Q: "Switch off the lights and then play some music"</span> <span class="s2"> A: ["Switch off the lights", "Play some music"]</span> <span class="s2"> Q: "I am feeling sad today, play some music"</span> <span class="s2"> A: ["Play some cheerful music"]</span> <span class="s2"> Q: "</span><span class="si">{</span><span class="n">command_text</span><span class="si">}</span><span class="s2">"</span> <span class="s2"> A: </span> <span class="s2"> """</span> <span class="n">completion</span> <span class="o">=</span> <span class="n">openai</span><span class="o">.</span><span class="n">Completion</span><span class="o">.</span><span class="n">create</span><span class="p">(</span><span class="n">engine</span><span class="o">=</span><span class="n">engine</span><span class="p">,</span> <span class="n">prompt</span><span class="o">=</span><span class="n">llm_prompt</span><span class="p">,</span> <span class="n">max_tokens</span><span class="o">=</span><span class="nb">len</span><span class="p">(</span><span class="n">command_text</span><span class="o">.</span><span class="n">split</span><span class="p">(</span><span class="s2">" "</span><span class="p">))</span><span class="o">*</span><span class="mi">2</span><span class="p">)</span> <span class="k">for</span> <span class="n">task</span> <span class="ow">in</span> <span class="nb">eval</span><span class="p">(</span><span class="n">completion</span><span class="o">.</span><span class="n">choices</span><span class="p">[</span><span class="mi">0</span><span class="p">]</span><span class="o">.</span><span class="n">text</span><span class="p">):</span> <span class="c1"># NEVER EVAL IN PROD RIGHT LIKE THIS</span> <span class="n">execute_command</span><span class="p">(</span><span class="n">task</span><span class="p">)</span> </code></pre> </div> <h3>Gluing together code</h3> <p>To finish it all off, we can use argparse to only send the input command to OpenAI when asked to do so.</p> <div class="codehilite"> <pre><span></span><code><span class="kn">import</span> <span class="nn">argparse</span> <span class="k">if</span> <span class="vm">__name__</span> <span class="o">==</span> <span class="s2">"__main__"</span><span class="p">:</span> <span class="n">parser</span> <span class="o">=</span> <span class="n">argparse</span><span class="o">.</span><span class="n">ArgumentParser</span><span class="p">()</span> <span class="n">parser</span><span class="o">.</span><span class="n">add_argument</span><span class="p">(</span><span class="s2">"command"</span><span class="p">,</span> <span class="n">nargs</span><span class="o">=</span><span class="s2">"?"</span><span class="p">,</span> <span class="nb">type</span><span class="o">=</span><span class="nb">str</span><span class="p">,</span> <span class="n">help</span><span class="o">=</span><span class="s2">"The command to pass to Siri"</span><span class="p">,</span> <span class="n">default</span><span class="o">=</span><span class="s2">"What time is it?"</span><span class="p">)</span> <span class="n">parser</span><span class="o">.</span><span class="n">add_argument</span><span class="p">(</span><span class="s1">'--openai'</span><span class="p">,</span> <span class="n">action</span><span class="o">=</span><span class="n">argparse</span><span class="o">.</span><span class="n">BooleanOptionalAction</span><span class="p">,</span> <span class="n">help</span><span class="o">=</span><span class="s2">"Use OpenAI to detect multiple intents"</span><span class="p">,</span> <span class="n">default</span><span class="o">=</span><span class="kc">False</span><span class="p">)</span> <span class="n">args</span> <span class="o">=</span> <span class="n">parser</span><span class="o">.</span><span class="n">parse_args</span><span class="p">()</span> <span class="k">if</span> <span class="n">args</span><span class="o">.</span><span class="n">openai</span><span class="p">:</span> <span class="n">execute_with_llm</span><span class="p">(</span><span class="n">args</span><span class="o">.</span><span class="n">command</span><span class="p">)</span> <span class="k">else</span><span class="p">:</span> <span class="n">execute_command</span><span class="p">(</span><span class="n">args</span><span class="o">.</span><span class="n">command</span><span class="p">)</span> </code></pre> </div> <h2>Conclusion</h2> <p>Siri is still dumb. When I ask it to <code>Switch off the lights</code>, it default to the home thousands of miles away. But, this code snippet definitely does work!</p> <blockquote>If you have scrolled this far, consider subscribing to my mailing list <a href="https://listmonk.navan.dev/subscription/form">here.</a> You can subscribe to either a specific type of post you are interested in, or subscribe to everything with the "Everything" list.</blockquote> <script data-isso="https://comments.navan.dev/" src="https://comments.navan.dev/js/embed.min.js"></script> <section id="isso-thread"> <noscript>Javascript needs to be activated to view comments.</noscript> </section> </main> <script src="assets/manup.min.js"></script> <script src="/pwabuilder-sw-register.js"></script> </body> </html>