an open-source alternative to the dots, bots and muses of the world, run an office of claude code/codex like agents on your laptop, sandboxes or anywhere, uses your existing subscriptions
HTML agent-harnessagent-orchestrationagentsai-agentsautonomous-agentsclaude-codecodexcoding-agentsdesktop-appdeveloper-toolselectrongemini-cliharnessharness-engineeringlocal-firstmemorymulti-agentopencodeorchestrationtypescript
docs(0.5.6): README and guides match the release; Pro is free until 24 October
Chaitanya Giri31bcf779d1f6
diff --git a/README.md b/README.mdindex e94e2582..b8cb8066 100644--- a/README.md+++ b/README.md@@ -59,8 +59,9 @@ meetings, all running locally.**New here?** Start with the [install guide](https://munderdiffl.in/blog/how-to-install-and-use-munder-difflin/), then[your first hour](https://munderdiffl.in/blog/your-first-hour-with-munder-difflin/), and see [every agent on our own floor](https://munderdiffl.in/blog/ai-agent-use-cases-our-office/).-<sub><b>Launch offer:</b> the Pro annual plan for $150 USD, adjusted for purchasing power in different-countries and as low as $100 a year. <a href="https://app.harnessmd.com/console/license">Get Pro</a></sub>+<sub><b>Launch offer:</b> Pro is free for everyone until Saturday 24 October 2026: download the app and sign in.+After that the Pro annual plan is $150 USD, adjusted for purchasing power in different countries and as low as $100 a+year. <a href="https://app.harnessmd.com/console/license">Get Pro</a></sub></div>@@ -321,18 +322,32 @@ than sending you to a docs page.- **Prerequisites**: one Settings page showing which supporting tools (uv, git, Node, MemPalace, each agent CLI) you have, what each is for, and a button that asks Michael to install what is missing.> [!NOTE]-> **Status: v0.5.6. Fewer tokens, temps for one job, and one Slack bot for every channel.**+> **Status: v0.5.6. Capabilities, Integrations and Teams, and fewer tokens.**+> **Pro is free for everyone until Saturday 24 October 2026**, as part of the launch: download the app and sign in.> These notes describe the 0.5.6 app on the [download page](https://harnessmd.com/download).+> **Capabilities (Pro)**: a marketplace of 2,347 MCP servers, skills, plugins and tools from the public list+> [HarnessMD/awesome-agent-capabilities](https://github.com/HarnessMD/awesome-agent-capabilities). We reviewed 622 of+> them; the rest read "Listed, not reviewed by us" and have no Install button. A merged change reaches the app with no+> new release, and the app uses the list only when it carries our signature.+> **Integrations**: one place for Slack, webhooks, the apps your agents use and your keys, each with a plain status and+> the last 50 things that happened. Nothing saves until you press Save.+> **Teams**: your office talks to your teammates' offices. One plain chat per teammate, you choose who answers each one,+> and files and folders go as a link with a password, for one download.> **Fewer tokens by default**: Compact when context fills and Clear on a switch of work are on for every new install.> The clear runs when an agent's work switches after a session of 60 minutes, and a `handoff.md` in the agent's folder-> carries the thread across. A new install has no hourly standup.+> carries the thread across. A new install has no hourly standup. In a replay of one office of 13 Claude agents over a+> week, with those two changes, token use fell about 37%. That is a replay, not a live measurement.+> **Agent files (Pro)**: export an agent as a file and choose what goes in; a key, a token or a secret's value never+> does. 36 templates replace role bundles.> **Temps**: any lasting agent can start a temp, a short lived agent for one job, not only the orchestrator. Wherever you-> pick who answers (Slack, a webhook, a schedule's Sends to, the Stapler's Send captures to), you can pick **A temp** or a-> temp that clones one of your agents. **Temps per agent** (0 to 4) and **Max temps at once** sit on the orchestrator's-> **Budget & breaker** tab.-> **Slack**: one bot can listen to **Every channel the bot is in**, a direct message to the bot is always answered, and-> replies stay in the thread. The Slack app needs four more scopes: `im:history`, `im:read`, `mpim:history`, `mpim:read`.-> Slack comes back by itself after a failed start.+> pick who answers (Slack, a teammate, a webhook, a schedule's Sends to, the Stapler's Send captures to), you can pick+> **A temp** or **A temp, clone of** one of your agents. **Temps per agent** (0 to 4) and **Max temps at once** sit on+> the orchestrator's **Budget & breaker** tab.+> **Slack**: one bot answers in **All channels** it is in by default, or only the channels you tick, and in direct+> messages; replies stay in the thread. Set the Slack app up from the manifest: press **Copy manifest** under+> **How to get these**. Slack comes back by itself after a failed start.+> **Token limits per period**: Every 5 hours, Daily, Weekly or Monthly, for the office and for each agent. They replace+> the old token caps.> **Sign in and accounts in the app**: press **Open browser to sign in** and sign in there; the screen moves on by itself,> or you paste the code into **Code from the browser**. **Sign out** and **Switch to Teams** sit in the profile menu.> **Circuit breaker**: an agent that stopped as told is no longer constrained again, and the same warning is not sent twice@@ -376,8 +391,6 @@ than sending you to a docs page.> renders one sidebar row with the task ledger read once.> Not in this release: the Kimi scroll and queue bug, which we could not reproduce, and it is first> on the 0.5.4 list.-> **Launch offer**: the Pro annual plan for $150 USD, adjusted for purchasing power in different-> countries and as low as $100 a year. [Get Pro](https://app.harnessmd.com/console/license)> Before that, v0.4.6 brought a Simplified Chinese and Arabic interface with right to left support> and self hosted fonts, the update badge that runs the real download and restart, and the security> work on CLI name validation and analytics.@@ -404,8 +417,9 @@ You do still need at least one agent CLI on your machine. From 0.5.3 a missing Cin the agent's terminal: Install runs the command, you sign in inside the app, and the agent starts.**Settings → Prerequisites** lists the rest.-**Pro** adds Stapler and the new sidebar. Launch offer: the annual plan-for $150 USD, adjusted for purchasing power, as low as $100 a year.+**Pro** adds Capabilities, agent files, Stapler and the new sidebar. As part of the launch it is free for everyone+until Saturday 24 October 2026. After that the annual plan is $150 USD, adjusted for purchasing power, as low as+$100 a year.[Get Pro](https://app.harnessmd.com/console/license)### Build from source@@ -468,7 +482,8 @@ visual system.## Roadmap-Shipped through **v0.5.6**: fewer tokens by default with a handoff file across a cleared session, temps that any+Shipped through **v0.5.6**: Capabilities, Integrations and Teams, agent files with 36 templates, token limits per+period, fewer tokens by default with a handoff file across a cleared session, temps that anyagent can start and that can answer inbound messages, one Slack bot for every channel it is in, sign in and accountsinside the app, a quieter circuit breaker, local dictation into any app and meetings with both sides of the call,with the model in the installer, the new agent cards in the free build and a new Pro sidebar, adiff --git a/blog/src/posts/ai-agent-use-cases-our-office.md b/blog/src/posts/ai-agent-use-cases-our-office.mdindex 71391b41..62dae329 100644--- a/blog/src/posts/ai-agent-use-cases-our-office.md+++ b/blog/src/posts/ai-agent-use-cases-our-office.md@@ -126,7 +126,7 @@ Not every job deserves a desk.Through files, on purpose. Every agent has its own folder with an inbox, an outbox and a memory file, and no agent writes in another agent's folder.-* **Messages.** An agent drops a small file in its outbox and the app delivers it. Two kinds of message, inform and done, end a thread, so two agents cannot reply to each other forever.+* **Messages.** An agent drops a small file in its outbox and the app delivers it. Two kinds of message, inform and done, end a thread, so two agents cannot reply to each other forever. A third kind, note, is information only and wakes nobody.* **The plan.** One written plan that only the orchestrator edits, and one task board of cards: todo, doing, blocked, done.* **The circuit breaker.** It watches every agent for loops, error storms and overspending. It steers first, then limits the agent to reading, then stops it.* **Secrets.** Keys never reach an agent. A local broker makes the outside call for it.diff --git a/blog/src/posts/how-to-install-and-use-munder-difflin.md b/blog/src/posts/how-to-install-and-use-munder-difflin.mdindex 6615465d..8600db9c 100644--- a/blog/src/posts/how-to-install-and-use-munder-difflin.md+++ b/blog/src/posts/how-to-install-and-use-munder-difflin.md@@ -27,12 +27,12 @@ faq:- q: "Can these agents really change files on my computer?"a: "Yes, and that is the point, so understand it before you start. An agent starts in the folder you give it and can read files, write files and run commands. It is not locked in there: it can reach anything your account can. Start with a new empty folder and watch what the agents do before you point one at anything you care about."- q: "Is the free version a separate download from Pro?"- a: "No. There is one installer. The free version is the Classic view, and a licence unlocks the Pro view in the same app. On first launch, press Continue with the free version to reach Classic."+ a: "No. There is one installer. The free version is the Classic view, and a licence unlocks the Pro view in the same app. On first launch, press Continue with the free version to reach Classic. Until Saturday 24 October 2026, Pro is free for everyone as part of the launch: sign in and the app opens Pro with no licence."startHere: truestartOrder: 1----<div class="callout tldr"><span class="ic">TL;DR</span><p>Pick one AI engine, by what you already pay for. Download Munder Difflin 0.5.6 and open it. Press <strong>Open browser to sign in</strong> and sign in, name your workspace and your orchestrator, then press <strong>Continue with the free version</strong>. If the engine is missing, the app offers to install it. About twenty minutes.</p></div>+<div class="callout tldr"><span class="ic">TL;DR</span><p>Pick one AI engine, by what you already pay for. Download Munder Difflin 0.5.6 and open it. Press <strong>Open browser to sign in</strong> and sign in, name your workspace and your orchestrator, then press <strong>Continue with the free version</strong>. Until Saturday 24 October 2026, Pro is free for everyone as part of the launch, so you go straight to Pro. If the engine is missing, the app offers to install it. About twenty minutes.</p></div>This is how to install Munder Difflin 0.5.6 when you have never used an AI coding tool. Most guides start three steps ahead of you. This one starts with what the thing is, and ends with a working office and a fix for every way setup can fail.@@ -190,6 +190,8 @@ Setup has four steps. The words in bold are the words on the screen.<figure class="mg" data-scene="signin"><img src="/blog/assets/media/how-to-install-and-use-munder-difflin/signin.png" width="1600" height="1200" loading="lazy" decoding="async" alt="Animation. Two windows on white. An app window on the left holds a button reading Open browser to sign in, an empty field labelled Code from the browser and a pale button reading Use this code. A dotted outline on the right marks the browser, which is not open yet. The first button is pressed and a browser window opens in the outline, with a crowded row of tabs, a tick, the words copy the code and a code drawn as a row of boxes with a dot in each. A copy of the code lifts out of the browser, travels across and drops into the field. Use this code turns yellow and is pressed, and the app window shows You are signed in."><figcaption>Sign in takes two windows: press Open browser to sign in and sign in there. If the screen does not move on by itself, copy the code from the browser, paste it into Code from the browser and press Use this code. Tap the app window to watch it again.</figcaption></figure>+Until Saturday 24 October 2026, Pro is free for everyone as part of the launch. Once you sign in, the app opens the Pro view with no licence and skips the screen below. From Sunday 25 October it is back.+Then comes the screen that stops people: the Pro sales screen, **Run your agents like a company**, with three buttons.* **Continue with the free version** opens the Classic view. This is the one you want if you have no licence.@@ -236,6 +238,8 @@ Almost everything. Setup asks for the minimum, and the rest lives in **Settings** **Memory & Knowledge** covers what agents remember. Search by meaning needs `uv` and MemPalace.* **Keys & Secrets** is where saved keys live. OpenCode, Crush, Pi and Qwen take a key; Claude Code and Codex use their own sign in.+In Pro, the **Integrations** row in the sidebar puts Slack, webhooks, the apps your agents use and your keys in one place, each with a plain status and the last 50 things that happened. Nothing saves until you press **Save**.+## What do you do when an install fails?Read the message first. The app says what is wrong in a sentence, and most screens carry the button that fixes it.diff --git a/blog/src/posts/your-first-hour-with-munder-difflin.md b/blog/src/posts/your-first-hour-with-munder-difflin.mdindex ddaf62bd..92377e67 100644--- a/blog/src/posts/your-first-hour-with-munder-difflin.md+++ b/blog/src/posts/your-first-hour-with-munder-difflin.md@@ -19,11 +19,11 @@ faq:- q: "How long does it take to get Munder Difflin running?"a: "Setup is four short steps when an AI engine is already installed and signed in: download, sign in, name your workspace and your orchestrator. The rest of the hour goes on a first job, a small team and the first work that starts without you."- q: "Do I need Pro for this first hour?"- a: "No. Agents, schedules, the context rules, webhooks, Slack and temps are in both views. The Pro view adds one window with screens for Tasks, Inbox, Automations, Memory and Capabilities, plus the Stapler, a small window above every app."+ a: "No. Agents, schedules, the context rules, webhooks, Slack and temps are in both views. The Pro view adds one window with screens for Tasks, Inbox, Automations, Memory and Capabilities, plus the Stapler, a small window above every app. Until Saturday 24 October 2026, Pro is free for everyone as part of the launch."- q: "What is a temp?"a: "A temp is a short lived agent started for one job. The orchestrator or any lasting agent can start one, and the app can start one to answer an inbound message. It does the job, reports done and leaves."- q: "Can one office mix Claude Code, Codex and other engines?"- a: "Yes. Each agent picks its own engine and model when you add it, and can have its own token cap. Claude Code and Codex use their own sign in; OpenCode, Crush, Pi and Qwen take a key."+ a: "Yes. Each agent picks its own engine and model when you add it, and can have its own token limit per period. Claude Code and Codex use their own sign in; OpenCode, Crush, Pi and Qwen take a key."- q: "Does the office keep working when I close my laptop?"a: "No. Agents, schedules, Slack and webhooks all run on your own computer, so they stop when the app closes or the laptop sleeps."- q: "Why did my agent's session clear itself?"@@ -48,7 +48,7 @@ The first launch asks **How are you using Munder Difflin?** Pick **On my own**,2. Sign in in the browser. The screen moves on by itself; if it does not, paste the code into **Code from the browser** and press **Use this code**.3. **Name your workspace**, **Meet your orchestrator**, then **Open the workspace**.-Next comes the Pro sales screen, **Run your agents like a company**. Press **Continue with the free version** to reach the Classic view, or **I have a licence key** to turn on Pro. The steps through minute 45 work in both views; a few labels below are the Pro view's.+Next comes the Pro sales screen, **Run your agents like a company**. Until Saturday 24 October 2026, Pro is free for everyone as part of the launch: once you sign in, the app opens the Pro view with no licence and skips this screen. From Sunday 25 October, press **Continue with the free version** to reach the Classic view, or **I have a licence key** to turn on Pro. The steps through minute 45 work in both views; a few labels below are the Pro view's.## Minute 10: how do you give your orchestrator a first job?@@ -73,7 +73,7 @@ One split to try on day one:* a reviewer on **Codex**, reading every change the builder makes* a helper on a free engine such as Antigravity, for docs and tidying-In the Pro view, each agent also has a **Token cap: the session budget before a restart**. Leave it on **Workspace budget** unless one agent needs its own. Sign every engine in before you hire on it, or the new agent cannot start; a Claude Code, Codex or Cursor agent stops at **Sign in needed**.+Each agent can also have its own **token limit per period**: **Every 5 hours**, **Daily**, **Weekly** or **Monthly**, with a cap, in the agent's side panel under **Budget**. Leave it empty unless one agent needs its own. Sign every engine in before you hire on it, or the new agent cannot start; a Claude Code, Codex or Cursor agent stops at **Sign in needed**.Our own floor has eleven kinds of seat: an orchestrator, mission control, a community lead, growth, motion graphics, research, a founder's office, a scrum master, an engineering lead, builders, and a designer. You do not need eleven. You need a reviewer.@@ -83,13 +83,13 @@ Our own floor has eleven kinds of seat: an orchestrator, mission control, a commWith an automation. In the Pro view open **Automations** and press **New automation**; in the Classic view the same things live under Triggers. Two kinds start there, a schedule and a webhook, and Slack is a third way in.-**A schedule** sends a message to an agent on a clock. Choose **Schedule**, then fill in four things: a **Label**, **When** (an **Interval** or **Weekly**), **Sends to**, and the **Message**. The message is what the agent should do when it fires. A good first one: every weekday morning, ask an agent to check something and write you three lines.+**A schedule** sends a message to an agent on a clock. Choose **Schedule**, then fill in four things: a **Label**, **When** (an **Interval** or **Weekly**), **Sends to**, and the **Message**, then press **Save**. The message is what the agent should do when it fires. A good first one: every weekday morning, ask an agent to check something and write you three lines.<figure class="mg" data-scene="schedule"><img src="/blog/assets/media/your-first-hour-with-munder-difflin/schedule.png" width="1600" height="1200" loading="lazy" decoding="async" alt="Animation. A card on white titled Automations with a yellow New automation button; on the right, a horizon with an empty sky and an office chair. A pointer presses New automation, picks Schedule over Webhook endpoint, and four fields fill in: Label, morning check; When, Weekly rather than Interval; Sends to, an agent; Message, check something and write you three lines. The pointer leaves. A sun climbs over the horizon, the card fires by itself, and a paper slip flies to the chair and gets a tick. The pointer peeks back in to check, then leaves again."><figcaption>A schedule has four parts: a Label, When (an Interval or Weekly), Sends to, and the Message the agent gets every time it fires. Tap the card to fire it again.</figcaption></figure>-**A webhook** lets any tool that can send a web request start work. Choose **Webhook endpoint**. A new endpoint starts switched off: choose how much you trust the caller, copy its address and its secret into the tool that will send, then turn it on. The sender must hold the secret.+**A webhook** lets any tool that can send a web request start work. Choose **Webhook endpoint**. A new endpoint starts switched off: choose how much you trust the caller, copy its address and its secret into the tool that will send, then turn it on and press **Save**. The sender must hold the secret.-**Slack** lets people send work from a channel and read the answer in the same thread. It is under **Settings**, then **Connections**. In 0.5.6 one bot can listen to **Every channel the bot is in**, and a direct message to the bot is always answered. For direct messages, add `im:history`, `im:read`, `mpim:history` and `mpim:read` on your Slack app's **OAuth & Permissions** page at api.slack.com, then follow the app's own **How to get these** steps for the events. You choose the agent under **Who answers inbound messages**. If Slack fails to start, it now comes back by itself. The [Slack guide](/blog/connect-slack-to-munder-difflin/) has the full setup.+**Slack** lets people send work from a channel or a direct message and read the answer in the same thread. In the Pro view it is under **Integrations**; in Classic, under **Settings**, then **Connections**. Create the Slack app from a manifest: press **Copy manifest** under **How to get these**, then in Slack choose **Create New App**, **From a manifest**. The channel list opens on **All channels**, every channel the bot is in, and direct messages to the bot are answered too. Tick channels only to narrow it. You choose the agent under **Who answers inbound messages**. If Slack fails to start, it now comes back by itself. The [Slack guide](/blog/connect-slack-to-munder-difflin/) has the full setup.<figure class="mg" data-scene="slack"><img src="/blog/assets/media/your-first-hour-with-munder-difflin/slack.png" width="1600" height="1200" loading="lazy" decoding="async" alt="Animation. On white, a Connections card under Settings with a Slack switch and a box for Every channel the bot is in; below it, three chat cards labelled a channel, another channel and a direct message. A pointer switches Slack on and ticks the box, and an office chair rolls into the empty seat under Who answers inbound messages. A string runs from each card to the chair. Each card posts a message, a note travels down its string to the chair, and a reply comes back and lands under that message as a thread. Then Slack switches itself off, the strings go slack, and it switches itself back on."><figcaption>In 0.5.6 one bot can hear every channel it is in, a direct message is always answered, and the reply comes back in the same thread. Tap a card to send a message.</figcaption></figure>@@ -100,7 +100,7 @@ A temp is an agent for one job. It gets an objective, does the work, reports donThree things changed in 0.5.6:* **Any lasting agent can start a temp**, not only the orchestrator. A temp cannot start a temp.-* **A temp can answer inbound messages.** Wherever you pick who answers (Slack, a webhook, a schedule's **Sends to**, the Stapler's **Send captures to**), the list now offers **A temp** and a temp that is a clone of one of your agents.+* **A temp can answer inbound messages.** Wherever you pick who answers (Slack, a teammate, a webhook, a schedule's **Sends to**, the Stapler's **Send captures to**), the list now offers **A temp** and **A temp, clone of** one of your agents.* **You set the limits.** In the Pro view, on the orchestrator's **Budget & breaker** tab, **Temps per agent** runs from 0 to 4 and **Max temps at once** caps how many temps run in the whole office at the same time. At 0, agents cannot start temps.<figure class="mg" data-scene="temp"><img src="/blog/assets/media/your-first-hour-with-munder-difflin/temp.png" width="1600" height="1200" loading="lazy" decoding="async" alt="Animation. Three office chairs stand on white over the words your lasting agents, beside the dotted outline of a folding chair. A message card reading one job waits at the top left, then flies across while a yellow folding chair slides in folded, half opens, snaps shut, then opens properly and gets a tag reading A temp. The message lands on its seat, a done card pops up and flies to the first office chair, and the folding chair folds flat and slides away. The three office chairs never move."><figcaption>A temp answers one message or does one job, reports done and leaves; your lasting agents keep their seats. Tap the message to send another.</figcaption></figure>@@ -135,7 +135,7 @@ One window with a sidebar, and a screen for each part of the office. The switch**Memory** lets you **Search every agent's memory** at once and see the result as a **Graph** or a **List**. Use it before you brief: if an agent already solved this last week, the note is here. Narrow a search with **Whose memory** and **Kind**.-**Capabilities** shows what each agent can reach: MCP servers, skills and tools.+**Capabilities** is a marketplace of 2,347 MCP servers, skills, plugins and tools from the public list [HarnessMD/awesome-agent-capabilities](https://github.com/HarnessMD/awesome-agent-capabilities). We reviewed 622 of them; the rest read **Listed, not reviewed by us** and have no **Install** button. A merged change to the list reaches the app with no new release.**The Stapler** is a small window that floats above every app. On the **Stapler** screen, turn on **Show the stapler**, then click it for its ring:@@ -152,7 +152,7 @@ The habit: when you see a bug in any app, do not describe it. Staple it.## Minute 58: what about accounts and teams?-Your account sits in the profile menu. **Sign out** asks you to press again, so you cannot do it by accident. **Switch to Teams** opens one sheet for joining a team with an invite code or generating one.+Your account sits in the profile menu. **Sign out** asks you to press again, so you cannot do it by accident. **Switch to Teams** opens one sheet for joining a team with an invite code or generating one. In 0.5.6 your office talks to your teammates' offices: one plain chat per teammate, you choose who answers each one, and files and folders go as a link with a password, for one download.## What trips people up in the first hour?diff --git a/docs/blog/ai-agent-use-cases-our-office/index.html b/docs/blog/ai-agent-use-cases-our-office/index.htmlindex 086dfa0b..8e3a7ae3 100644--- a/docs/blog/ai-agent-use-cases-our-office/index.html+++ b/docs/blog/ai-agent-use-cases-our-office/index.html@@ -322,7 +322,7 @@<h2 id="how-does-work-move-between-them" tabindex="-1">How does work move between them? <a class="anchor" href="#how-does-work-move-between-them" aria-hidden="true">#</a></h2><p>Through files, on purpose. Every agent has its own folder with an inbox, an outbox and a memory file, and no agent writes in another agent’s folder.</p><ul>-<li><strong>Messages.</strong> An agent drops a small file in its outbox and the app delivers it. Two kinds of message, inform and done, end a thread, so two agents cannot reply to each other forever.</li>+<li><strong>Messages.</strong> An agent drops a small file in its outbox and the app delivers it. Two kinds of message, inform and done, end a thread, so two agents cannot reply to each other forever. A third kind, note, is information only and wakes nobody.</li><li><strong>The plan.</strong> One written plan that only the orchestrator edits, and one task board of cards: todo, doing, blocked, done.</li><li><strong>The circuit breaker.</strong> It watches every agent for loops, error storms and overspending. It steers first, then limits the agent to reading, then stops it.</li><li><strong>Secrets.</strong> Keys never reach an agent. A local broker makes the outside call for it.</li>diff --git a/docs/blog/how-to-install-and-use-munder-difflin/index.html b/docs/blog/how-to-install-and-use-munder-difflin/index.htmlindex 34bb5bf3..23cf24e7 100644--- a/docs/blog/how-to-install-and-use-munder-difflin/index.html+++ b/docs/blog/how-to-install-and-use-munder-difflin/index.html@@ -102,7 +102,7 @@{ "@type": "Question", "name": "Do I need an account?", "acceptedAnswer": { "@type": "Answer", "text": "Yes. The first launch asks you to sign in through your browser, on the free version as well. The browser does not open by itself: press Open browser to sign in and sign in there. The screen usually moves on by itself; if it does not, paste the code the browser shows you into the app." } },{ "@type": "Question", "name": "Which operating systems does Munder Difflin run on?", "acceptedAnswer": { "@type": "Answer", "text": "macOS, Windows and Linux. macOS is one universal build for Apple Silicon and Intel, signed and notarized by Apple. Windows 10 and 11, 64 bit, get a setup installer and a portable build. Linux gets a 64 bit AppImage." } },{ "@type": "Question", "name": "Can these agents really change files on my computer?", "acceptedAnswer": { "@type": "Answer", "text": "Yes, and that is the point, so understand it before you start. An agent starts in the folder you give it and can read files, write files and run commands. It is not locked in there: it can reach anything your account can. Start with a new empty folder and watch what the agents do before you point one at anything you care about." } },- { "@type": "Question", "name": "Is the free version a separate download from Pro?", "acceptedAnswer": { "@type": "Answer", "text": "No. There is one installer. The free version is the Classic view, and a licence unlocks the Pro view in the same app. On first launch, press Continue with the free version to reach Classic." } }+ { "@type": "Question", "name": "Is the free version a separate download from Pro?", "acceptedAnswer": { "@type": "Answer", "text": "No. There is one installer. The free version is the Classic view, and a licence unlocks the Pro view in the same app. On first launch, press Continue with the free version to reach Classic. Until Saturday 24 October 2026, Pro is free for everyone as part of the launch: sign in and the app opens Pro with no licence." } }]}@@ -188,7 +188,7 @@<div class="post-wrap wrap"><div class="prose">- <div class="callout tldr"><span class="ic">TL;DR</span><p>Pick one AI engine, by what you already pay for. Download Munder Difflin 0.5.6 and open it. Press <strong>Open browser to sign in</strong> and sign in, name your workspace and your orchestrator, then press <strong>Continue with the free version</strong>. If the engine is missing, the app offers to install it. About twenty minutes.</p></div>+ <div class="callout tldr"><span class="ic">TL;DR</span><p>Pick one AI engine, by what you already pay for. Download Munder Difflin 0.5.6 and open it. Press <strong>Open browser to sign in</strong> and sign in, name your workspace and your orchestrator, then press <strong>Continue with the free version</strong>. Until Saturday 24 October 2026, Pro is free for everyone as part of the launch, so you go straight to Pro. If the engine is missing, the app offers to install it. About twenty minutes.</p></div><p>This is how to install Munder Difflin 0.5.6 when you have never used an AI coding tool. Most guides start three steps ahead of you. This one starts with what the thing is, and ends with a working office and a fix for every way setup can fail.</p><h2 id="what-is-munder-difflin" tabindex="-1">What is Munder Difflin? <a class="anchor" href="#what-is-munder-difflin" aria-hidden="true">#</a></h2><p><a href="https://harnessmd.com/download">Munder Difflin</a> is free and open source: a desktop app that runs a team of coding agents such as Claude Code, Codex and Gemini CLI on your own computer.</p>@@ -324,6 +324,7 @@ opencode auth login</code></pre><li><strong>A last look at him.</strong> Press <strong>Open the workspace</strong>, and you are in.</li></ol><figure class="mg" data-scene="signin"><img src="/blog/assets/media/how-to-install-and-use-munder-difflin/signin.png" width="1600" height="1200" loading="lazy" decoding="async" alt="Animation. Two windows on white. An app window on the left holds a button reading Open browser to sign in, an empty field labelled Code from the browser and a pale button reading Use this code. A dotted outline on the right marks the browser, which is not open yet. The first button is pressed and a browser window opens in the outline, with a crowded row of tabs, a tick, the words copy the code and a code drawn as a row of boxes with a dot in each. A copy of the code lifts out of the browser, travels across and drops into the field. Use this code turns yellow and is pressed, and the app window shows You are signed in."><figcaption>Sign in takes two windows: press Open browser to sign in and sign in there. If the screen does not move on by itself, copy the code from the browser, paste it into Code from the browser and press Use this code. Tap the app window to watch it again.</figcaption></figure>+<p>Until Saturday 24 October 2026, Pro is free for everyone as part of the launch. Once you sign in, the app opens the Pro view with no licence and skips the screen below. From Sunday 25 October it is back.</p><p>Then comes the screen that stops people: the Pro sales screen, <strong>Run your agents like a company</strong>, with three buttons.</p><ul><li><strong>Continue with the free version</strong> opens the Classic view. This is the one you want if you have no licence.</li>@@ -358,6 +359,7 @@ opencode auth login</code></pre><li><strong>Memory & Knowledge</strong> covers what agents remember. Search by meaning needs <code>uv</code> and MemPalace.</li><li><strong>Keys & Secrets</strong> is where saved keys live. OpenCode, Crush, Pi and Qwen take a key; Claude Code and Codex use their own sign in.</li></ul>+<p>In Pro, the <strong>Integrations</strong> row in the sidebar puts Slack, webhooks, the apps your agents use and your keys in one place, each with a plain status and the last 50 things that happened. Nothing saves until you press <strong>Save</strong>.</p><h2 id="what-do-you-do-when-an-install-fails" tabindex="-1">What do you do when an install fails? <a class="anchor" href="#what-do-you-do-when-an-install-fails" aria-hidden="true">#</a></h2><p>Read the message first. The app says what is wrong in a sentence, and most screens carry the button that fixes it.</p><div class="table-scroll">@@ -460,7 +462,7 @@ opencode auth login</code></pre><div class="qa"><p class="q">Is the free version a separate download from Pro?</p>- <p class="a">No. There is one installer. The free version is the Classic view, and a licence unlocks the Pro view in the same app. On first launch, press Continue with the free version to reach Classic.</p>+ <p class="a">No. There is one installer. The free version is the Classic view, and a licence unlocks the Pro view in the same app. On first launch, press Continue with the free version to reach Classic. Until Saturday 24 October 2026, Pro is free for everyone as part of the launch: sign in and the app opens Pro with no licence.</p></div></section>diff --git a/docs/blog/index.html b/docs/blog/index.htmlindex 9cdca5a4..224f0729 100644--- a/docs/blog/index.html+++ b/docs/blog/index.html@@ -197,7 +197,7 @@<div class="meta"><time datetime="2026-07-03">Jul 3, 2026</time><span class="dot" aria-hidden="true"></span>- <span>10 min</span><span class="pin">Pinned</span>+ <span>11 min</span><span class="pin">Pinned</span></div><h3><a href="/blog/your-first-hour-with-munder-difflin/">Your First Hour With Munder Difflin 0.5.6: Onboarding From First Sign In to a Working Office</a></h3><p class="dek">Your first hour with Munder Difflin 0.5.6: sign in, brief your orchestrator, hire a team, schedule work, connect Slack, start temps, then tour the Pro screens.</p>diff --git a/docs/blog/tags/getting-started/index.html b/docs/blog/tags/getting-started/index.htmlindex 182d8ccc..925dd56e 100644--- a/docs/blog/tags/getting-started/index.html+++ b/docs/blog/tags/getting-started/index.html@@ -307,7 +307,7 @@<div class="meta"><time datetime="2026-07-03">Jul 3, 2026</time><span class="dot" aria-hidden="true"></span>- <span>10 min</span><span class="pin">Pinned</span>+ <span>11 min</span><span class="pin">Pinned</span></div><h3><a href="/blog/your-first-hour-with-munder-difflin/">Your First Hour With Munder Difflin 0.5.6: Onboarding From First Sign In to a Working Office</a></h3><p class="dek">Your first hour with Munder Difflin 0.5.6: sign in, brief your orchestrator, hire a team, schedule work, connect Slack, start temps, then tour the Pro screens.</p>diff --git a/docs/blog/tags/guides/index.html b/docs/blog/tags/guides/index.htmlindex fd26f1ba..cab7b13e 100644--- a/docs/blog/tags/guides/index.html+++ b/docs/blog/tags/guides/index.html@@ -467,7 +467,7 @@<div class="meta"><time datetime="2026-07-03">Jul 3, 2026</time><span class="dot" aria-hidden="true"></span>- <span>10 min</span><span class="pin">Pinned</span>+ <span>11 min</span><span class="pin">Pinned</span></div><h3><a href="/blog/your-first-hour-with-munder-difflin/">Your First Hour With Munder Difflin 0.5.6: Onboarding From First Sign In to a Working Office</a></h3><p class="dek">Your first hour with Munder Difflin 0.5.6: sign in, brief your orchestrator, hire a team, schedule work, connect Slack, start temps, then tour the Pro screens.</p>diff --git a/docs/blog/tags/multi-agent/index.html b/docs/blog/tags/multi-agent/index.htmlindex eb5d6528..194d116a 100644--- a/docs/blog/tags/multi-agent/index.html+++ b/docs/blog/tags/multi-agent/index.html@@ -483,7 +483,7 @@<div class="meta"><time datetime="2026-07-03">Jul 3, 2026</time><span class="dot" aria-hidden="true"></span>- <span>10 min</span><span class="pin">Pinned</span>+ <span>11 min</span><span class="pin">Pinned</span></div><h3><a href="/blog/your-first-hour-with-munder-difflin/">Your First Hour With Munder Difflin 0.5.6: Onboarding From First Sign In to a Working Office</a></h3><p class="dek">Your first hour with Munder Difflin 0.5.6: sign in, brief your orchestrator, hire a team, schedule work, connect Slack, start temps, then tour the Pro screens.</p>diff --git a/docs/blog/tags/onboarding/index.html b/docs/blog/tags/onboarding/index.htmlindex 5000dbfc..17d3774a 100644--- a/docs/blog/tags/onboarding/index.html+++ b/docs/blog/tags/onboarding/index.html@@ -115,7 +115,7 @@<div class="meta"><time datetime="2026-07-03">Jul 3, 2026</time><span class="dot" aria-hidden="true"></span>- <span>10 min</span><span class="pin">Pinned</span>+ <span>11 min</span><span class="pin">Pinned</span></div><h3><a href="/blog/your-first-hour-with-munder-difflin/">Your First Hour With Munder Difflin 0.5.6: Onboarding From First Sign In to a Working Office</a></h3><p class="dek">Your first hour with Munder Difflin 0.5.6: sign in, brief your orchestrator, hire a team, schedule work, connect Slack, start temps, then tour the Pro screens.</p>diff --git a/docs/blog/topics/guides/index.html b/docs/blog/topics/guides/index.htmlindex fa57197e..79dd209d 100644--- a/docs/blog/topics/guides/index.html+++ b/docs/blog/topics/guides/index.html@@ -159,7 +159,7 @@<div class="meta"><time datetime="2026-07-03">Jul 3, 2026</time><span class="dot" aria-hidden="true"></span>- <span>10 min</span><span class="pin">Pinned</span>+ <span>11 min</span><span class="pin">Pinned</span></div><h3><a href="/blog/your-first-hour-with-munder-difflin/">Your First Hour With Munder Difflin 0.5.6: Onboarding From First Sign In to a Working Office</a></h3><p class="dek">Your first hour with Munder Difflin 0.5.6: sign in, brief your orchestrator, hire a team, schedule work, connect Slack, start temps, then tour the Pro screens.</p>diff --git a/docs/blog/your-first-hour-with-munder-difflin/index.html b/docs/blog/your-first-hour-with-munder-difflin/index.htmlindex a7f24dbe..1ec48550 100644--- a/docs/blog/your-first-hour-with-munder-difflin/index.html+++ b/docs/blog/your-first-hour-with-munder-difflin/index.html@@ -98,9 +98,9 @@"@type": "FAQPage","mainEntity": [{ "@type": "Question", "name": "How long does it take to get Munder Difflin running?", "acceptedAnswer": { "@type": "Answer", "text": "Setup is four short steps when an AI engine is already installed and signed in: download, sign in, name your workspace and your orchestrator. The rest of the hour goes on a first job, a small team and the first work that starts without you." } },- { "@type": "Question", "name": "Do I need Pro for this first hour?", "acceptedAnswer": { "@type": "Answer", "text": "No. Agents, schedules, the context rules, webhooks, Slack and temps are in both views. The Pro view adds one window with screens for Tasks, Inbox, Automations, Memory and Capabilities, plus the Stapler, a small window above every app." } },+ { "@type": "Question", "name": "Do I need Pro for this first hour?", "acceptedAnswer": { "@type": "Answer", "text": "No. Agents, schedules, the context rules, webhooks, Slack and temps are in both views. The Pro view adds one window with screens for Tasks, Inbox, Automations, Memory and Capabilities, plus the Stapler, a small window above every app. Until Saturday 24 October 2026, Pro is free for everyone as part of the launch." } },{ "@type": "Question", "name": "What is a temp?", "acceptedAnswer": { "@type": "Answer", "text": "A temp is a short lived agent started for one job. The orchestrator or any lasting agent can start one, and the app can start one to answer an inbound message. It does the job, reports done and leaves." } },- { "@type": "Question", "name": "Can one office mix Claude Code, Codex and other engines?", "acceptedAnswer": { "@type": "Answer", "text": "Yes. Each agent picks its own engine and model when you add it, and can have its own token cap. Claude Code and Codex use their own sign in; OpenCode, Crush, Pi and Qwen take a key." } },+ { "@type": "Question", "name": "Can one office mix Claude Code, Codex and other engines?", "acceptedAnswer": { "@type": "Answer", "text": "Yes. Each agent picks its own engine and model when you add it, and can have its own token limit per period. Claude Code and Codex use their own sign in; OpenCode, Crush, Pi and Qwen take a key." } },{ "@type": "Question", "name": "Does the office keep working when I close my laptop?", "acceptedAnswer": { "@type": "Answer", "text": "No. Agents, schedules, Slack and webhooks all run on your own computer, so they stop when the app closes or the laptop sleeps." } },{ "@type": "Question", "name": "Why did my agent's session clear itself?", "acceptedAnswer": { "@type": "Answer", "text": "That is a default in 0.5.6 that saves tokens. When an agent's work switches and its session has run for 60 minutes, the app clears the session, and a handoff file in the agent's folder carries the thread across." } }@@ -179,7 +179,7 @@<span class="dot" aria-hidden="true">·</span><time datetime="2026-07-03">Jul 3, 2026</time><span class="dot" aria-hidden="true">·</span>- <span>10 min read</span>+ <span>11 min read</span><span class="tag kind">Guides</span></div></header>@@ -199,7 +199,7 @@<li>Sign in in the browser. The screen moves on by itself; if it does not, paste the code into <strong>Code from the browser</strong> and press <strong>Use this code</strong>.</li><li><strong>Name your workspace</strong>, <strong>Meet your orchestrator</strong>, then <strong>Open the workspace</strong>.</li></ol>-<p>Next comes the Pro sales screen, <strong>Run your agents like a company</strong>. Press <strong>Continue with the free version</strong> to reach the Classic view, or <strong>I have a licence key</strong> to turn on Pro. The steps through minute 45 work in both views; a few labels below are the Pro view’s.</p>+<p>Next comes the Pro sales screen, <strong>Run your agents like a company</strong>. Until Saturday 24 October 2026, Pro is free for everyone as part of the launch: once you sign in, the app opens the Pro view with no licence and skips this screen. From Sunday 25 October, press <strong>Continue with the free version</strong> to reach the Classic view, or <strong>I have a licence key</strong> to turn on Pro. The steps through minute 45 work in both views; a few labels below are the Pro view’s.</p><h2 id="minute-10-how-do-you-give-your-orchestrator-a-first-job" tabindex="-1">Minute 10: how do you give your orchestrator a first job? <a class="anchor" href="#minute-10-how-do-you-give-your-orchestrator-a-first-job" aria-hidden="true">#</a></h2><p>Brief him like a capable new hire: the outcome, the folder, and what done looks like. Your orchestrator is the one agent you talk to. He is called Michael until you rename him.</p><p>A vague brief gets a vague job. Compare:</p>@@ -217,22 +217,22 @@<li>a reviewer on <strong>Codex</strong>, reading every change the builder makes</li><li>a helper on a free engine such as Antigravity, for docs and tidying</li></ul>-<p>In the Pro view, each agent also has a <strong>Token cap: the session budget before a restart</strong>. Leave it on <strong>Workspace budget</strong> unless one agent needs its own. Sign every engine in before you hire on it, or the new agent cannot start; a Claude Code, Codex or Cursor agent stops at <strong>Sign in needed</strong>.</p>+<p>Each agent can also have its own <strong>token limit per period</strong>: <strong>Every 5 hours</strong>, <strong>Daily</strong>, <strong>Weekly</strong> or <strong>Monthly</strong>, with a cap, in the agent’s side panel under <strong>Budget</strong>. Leave it empty unless one agent needs its own. Sign every engine in before you hire on it, or the new agent cannot start; a Claude Code, Codex or Cursor agent stops at <strong>Sign in needed</strong>.</p><p>Our own floor has eleven kinds of seat: an orchestrator, mission control, a community lead, growth, motion graphics, research, a founder’s office, a scrum master, an engineering lead, builders, and a designer. You do not need eleven. You need a reviewer.</p><figure class="mg" data-scene="team"><img src="/blog/assets/media/your-first-hour-with-munder-difflin/team.png" width="1600" height="1200" loading="lazy" decoding="async" alt="Animation. Three dotted outlines of office chairs stand in a row on white under a yellow button reading Add an agent, with the roles builder, reviewer and helper below them. A pointer presses the button three times; each press rolls in a chair, and an engine tag appears under it: on Claude Code, on Codex, on a free engine. The third chair arrives still spinning. A sheet of paper leaves the builder's chair, crosses to the reviewer's chair, is read line by line, gets a tick and comes back. Then the helper's heap of papers squares itself into a neat stack."><figcaption>A first team of three: a builder on Claude Code, a reviewer on Codex that reads every change, and a helper on a free engine for docs and tidying. Tap the builder to send another change for review.</figcaption></figure><h2 id="minute-35-how-do-you-set-up-work-that-starts-without-you" tabindex="-1">Minute 35: how do you set up work that starts without you? <a class="anchor" href="#minute-35-how-do-you-set-up-work-that-starts-without-you" aria-hidden="true">#</a></h2><p>With an automation. In the Pro view open <strong>Automations</strong> and press <strong>New automation</strong>; in the Classic view the same things live under Triggers. Two kinds start there, a schedule and a webhook, and Slack is a third way in.</p>-<p><strong>A schedule</strong> sends a message to an agent on a clock. Choose <strong>Schedule</strong>, then fill in four things: a <strong>Label</strong>, <strong>When</strong> (an <strong>Interval</strong> or <strong>Weekly</strong>), <strong>Sends to</strong>, and the <strong>Message</strong>. The message is what the agent should do when it fires. A good first one: every weekday morning, ask an agent to check something and write you three lines.</p>+<p><strong>A schedule</strong> sends a message to an agent on a clock. Choose <strong>Schedule</strong>, then fill in four things: a <strong>Label</strong>, <strong>When</strong> (an <strong>Interval</strong> or <strong>Weekly</strong>), <strong>Sends to</strong>, and the <strong>Message</strong>, then press <strong>Save</strong>. The message is what the agent should do when it fires. A good first one: every weekday morning, ask an agent to check something and write you three lines.</p><figure class="mg" data-scene="schedule"><img src="/blog/assets/media/your-first-hour-with-munder-difflin/schedule.png" width="1600" height="1200" loading="lazy" decoding="async" alt="Animation. A card on white titled Automations with a yellow New automation button; on the right, a horizon with an empty sky and an office chair. A pointer presses New automation, picks Schedule over Webhook endpoint, and four fields fill in: Label, morning check; When, Weekly rather than Interval; Sends to, an agent; Message, check something and write you three lines. The pointer leaves. A sun climbs over the horizon, the card fires by itself, and a paper slip flies to the chair and gets a tick. The pointer peeks back in to check, then leaves again."><figcaption>A schedule has four parts: a Label, When (an Interval or Weekly), Sends to, and the Message the agent gets every time it fires. Tap the card to fire it again.</figcaption></figure>-<p><strong>A webhook</strong> lets any tool that can send a web request start work. Choose <strong>Webhook endpoint</strong>. A new endpoint starts switched off: choose how much you trust the caller, copy its address and its secret into the tool that will send, then turn it on. The sender must hold the secret.</p>-<p><strong>Slack</strong> lets people send work from a channel and read the answer in the same thread. It is under <strong>Settings</strong>, then <strong>Connections</strong>. In 0.5.6 one bot can listen to <strong>Every channel the bot is in</strong>, and a direct message to the bot is always answered. For direct messages, add <code>im:history</code>, <code>im:read</code>, <code>mpim:history</code> and <code>mpim:read</code> on your Slack app’s <strong>OAuth & Permissions</strong> page at <a href="http://api.slack.com">api.slack.com</a>, then follow the app’s own <strong>How to get these</strong> steps for the events. You choose the agent under <strong>Who answers inbound messages</strong>. If Slack fails to start, it now comes back by itself. The <a href="/blog/connect-slack-to-munder-difflin/">Slack guide</a> has the full setup.</p>+<p><strong>A webhook</strong> lets any tool that can send a web request start work. Choose <strong>Webhook endpoint</strong>. A new endpoint starts switched off: choose how much you trust the caller, copy its address and its secret into the tool that will send, then turn it on and press <strong>Save</strong>. The sender must hold the secret.</p>+<p><strong>Slack</strong> lets people send work from a channel or a direct message and read the answer in the same thread. In the Pro view it is under <strong>Integrations</strong>; in Classic, under <strong>Settings</strong>, then <strong>Connections</strong>. Create the Slack app from a manifest: press <strong>Copy manifest</strong> under <strong>How to get these</strong>, then in Slack choose <strong>Create New App</strong>, <strong>From a manifest</strong>. The channel list opens on <strong>All channels</strong>, every channel the bot is in, and direct messages to the bot are answered too. Tick channels only to narrow it. You choose the agent under <strong>Who answers inbound messages</strong>. If Slack fails to start, it now comes back by itself. The <a href="/blog/connect-slack-to-munder-difflin/">Slack guide</a> has the full setup.</p><figure class="mg" data-scene="slack"><img src="/blog/assets/media/your-first-hour-with-munder-difflin/slack.png" width="1600" height="1200" loading="lazy" decoding="async" alt="Animation. On white, a Connections card under Settings with a Slack switch and a box for Every channel the bot is in; below it, three chat cards labelled a channel, another channel and a direct message. A pointer switches Slack on and ticks the box, and an office chair rolls into the empty seat under Who answers inbound messages. A string runs from each card to the chair. Each card posts a message, a note travels down its string to the chair, and a reply comes back and lands under that message as a thread. Then Slack switches itself off, the strings go slack, and it switches itself back on."><figcaption>In 0.5.6 one bot can hear every channel it is in, a direct message is always answered, and the reply comes back in the same thread. Tap a card to send a message.</figcaption></figure><h2 id="minute-45-what-are-temps-and-how-do-you-keep-token-use-down" tabindex="-1">Minute 45: what are temps, and how do you keep token use down? <a class="anchor" href="#minute-45-what-are-temps-and-how-do-you-keep-token-use-down" aria-hidden="true">#</a></h2><p>A temp is an agent for one job. It gets an objective, does the work, reports done and leaves. There is no button for one. You ask: <em>“Michael, start a temp to rename every file in this folder to lower case.”</em></p><p>Three things changed in 0.5.6:</p><ul><li><strong>Any lasting agent can start a temp</strong>, not only the orchestrator. A temp cannot start a temp.</li>-<li><strong>A temp can answer inbound messages.</strong> Wherever you pick who answers (Slack, a webhook, a schedule’s <strong>Sends to</strong>, the Stapler’s <strong>Send captures to</strong>), the list now offers <strong>A temp</strong> and a temp that is a clone of one of your agents.</li>+<li><strong>A temp can answer inbound messages.</strong> Wherever you pick who answers (Slack, a teammate, a webhook, a schedule’s <strong>Sends to</strong>, the Stapler’s <strong>Send captures to</strong>), the list now offers <strong>A temp</strong> and <strong>A temp, clone of</strong> one of your agents.</li><li><strong>You set the limits.</strong> In the Pro view, on the orchestrator’s <strong>Budget & breaker</strong> tab, <strong>Temps per agent</strong> runs from 0 to 4 and <strong>Max temps at once</strong> caps how many temps run in the whole office at the same time. At 0, agents cannot start temps.</li></ul><figure class="mg" data-scene="temp"><img src="/blog/assets/media/your-first-hour-with-munder-difflin/temp.png" width="1600" height="1200" loading="lazy" decoding="async" alt="Animation. Three office chairs stand on white over the words your lasting agents, beside the dotted outline of a folding chair. A message card reading one job waits at the top left, then flies across while a yellow folding chair slides in folded, half opens, snaps shut, then opens properly and gets a tag reading A temp. The message lands on its seat, a done card pops up and flies to the first office chair, and the folding chair folds flat and slides away. The three office chairs never move."><figcaption>A temp answers one message or does one job, reports done and leaves; your lasting agents keep their seats. Tap the message to send another.</figcaption></figure>@@ -254,7 +254,7 @@<figure class="mg" data-scene="askme"><img src="/blog/assets/media/your-first-hour-with-munder-difflin/askme.png" width="1600" height="1200" loading="lazy" decoding="async" alt="Animation. On white, a large card with an Ask me chip and the word Inbox shows a question as grey lines with a lilac question mark, an answer field reading Answer in one line, a blue send button and a chip reading waiting on you. Beside it, three smaller question cards wait in a column under the words waiting on a person. A short answer types itself in as one line and is sent; the question mark turns into a mint tick and the chip becomes unblocks the task. The card slides away, the next question card grows into its place, and the queue shuffles up; the last card arrives a little too fast and bumps the one ahead. The new question gets its one line answer and its tick too."><figcaption>Ask me collects the questions waiting on a person. A short answer unblocks the task, and the card moves on. Tap the card to ask again.</figcaption></figure><p><strong>Automations</strong> is the list you met at minute 35, with filters for <strong>Scheduled</strong>, <strong>Context</strong> and <strong>Webhooks</strong>, Each schedule shows when it <strong>Last fired</strong>, each webhook lists its <strong>Recent</strong> arrivals, and <strong>History</strong> opens the Inbox, where webhook messages are answered.</p><p><strong>Memory</strong> lets you <strong>Search every agent’s memory</strong> at once and see the result as a <strong>Graph</strong> or a <strong>List</strong>. Use it before you brief: if an agent already solved this last week, the note is here. Narrow a search with <strong>Whose memory</strong> and <strong>Kind</strong>.</p>-<p><strong>Capabilities</strong> shows what each agent can reach: MCP servers, skills and tools.</p>+<p><strong>Capabilities</strong> is a marketplace of 2,347 MCP servers, skills, plugins and tools from the public list <a href="https://github.com/HarnessMD/awesome-agent-capabilities">HarnessMD/awesome-agent-capabilities</a>. We reviewed 622 of them; the rest read <strong>Listed, not reviewed by us</strong> and have no <strong>Install</strong> button. A merged change to the list reaches the app with no new release.</p><p><strong>The Stapler</strong> is a small window that floats above every app. On the <strong>Stapler</strong> screen, turn on <strong>Show the stapler</strong>, then click it for its ring:</p><ul><li><strong>Screenshot</strong> captures a region of the screen and sends it, with a note, to the agent you pick under <strong>Send captures to</strong>.</li>@@ -266,7 +266,7 @@<p>The habit: when you see a bug in any app, do not describe it. Staple it.</p><figure class="mg" data-scene="stapler"><img src="/blog/assets/media/your-first-hour-with-munder-difflin/stapler.png" width="1600" height="1200" loading="lazy" decoding="async" alt="Animation. On white, a plain app window made of grey bars, with a setting above it reading Show the stapler and a chip reading an agent. The setting switches on and a round stapler button pops up and floats over the corner of the app window. A pointer clicks it and a ring of four actions opens around it: Screenshot, Record message, Record meeting, Make invisible. Screenshot is picked, a crosshair drags a dashed box over part of the app, and the region lifts out as a capture. A note field reading your note slides in under it and a short line types itself in. The stapler's arm comes down and a staple joins the note to the capture, then the pair flies to the agent chip, which gets a tick. The ring opens again and stays."><figcaption>The Stapler floats above every app. Its ring offers Screenshot, Record message, Record meeting and Make invisible, plus Computer use, coming soon, and a capture goes to an agent with your note. Tap the stapler to open the ring.</figcaption></figure><h2 id="minute-58-what-about-accounts-and-teams" tabindex="-1">Minute 58: what about accounts and teams? <a class="anchor" href="#minute-58-what-about-accounts-and-teams" aria-hidden="true">#</a></h2>-<p>Your account sits in the profile menu. <strong>Sign out</strong> asks you to press again, so you cannot do it by accident. <strong>Switch to Teams</strong> opens one sheet for joining a team with an invite code or generating one.</p>+<p>Your account sits in the profile menu. <strong>Sign out</strong> asks you to press again, so you cannot do it by accident. <strong>Switch to Teams</strong> opens one sheet for joining a team with an invite code or generating one. In 0.5.6 your office talks to your teammates’ offices: one plain chat per teammate, you choose who answers each one, and files and folders go as a link with a password, for one download.</p><h2 id="what-trips-people-up-in-the-first-hour" tabindex="-1">What trips people up in the first hour? <a class="anchor" href="#what-trips-people-up-in-the-first-hour" aria-hidden="true">#</a></h2><p>Eight things, all normal.</p><ol>@@ -295,7 +295,7 @@<div class="qa"><p class="q">Do I need Pro for this first hour?</p>- <p class="a">No. Agents, schedules, the context rules, webhooks, Slack and temps are in both views. The Pro view adds one window with screens for Tasks, Inbox, Automations, Memory and Capabilities, plus the Stapler, a small window above every app.</p>+ <p class="a">No. Agents, schedules, the context rules, webhooks, Slack and temps are in both views. The Pro view adds one window with screens for Tasks, Inbox, Automations, Memory and Capabilities, plus the Stapler, a small window above every app. Until Saturday 24 October 2026, Pro is free for everyone as part of the launch.</p></div><div class="qa">@@ -305,7 +305,7 @@<div class="qa"><p class="q">Can one office mix Claude Code, Codex and other engines?</p>- <p class="a">Yes. Each agent picks its own engine and model when you add it, and can have its own token cap. Claude Code and Codex use their own sign in; OpenCode, Crush, Pi and Qwen take a key.</p>+ <p class="a">Yes. Each agent picks its own engine and model when you add it, and can have its own token limit per period. Claude Code and Codex use their own sign in; OpenCode, Crush, Pi and Qwen take a key.</p></div><div class="qa">