# Exponential: Vibecode together — Issues, agents, one workspace

> Real-time AI team issue tracker, customer support, and coding agents in one workspace. (iPhone/iPad app by Dennis Strahhuber.)

- Source: https://appshunter.io/ios/app/exponential-vibecode-together/id6788189402 (this page in markdown: same URL + `.md`)
- Developer: [Dennis Strahhuber](https://appshunter.io/developer/6788189404)
- Category: Productivity, Developer Tools
- Price: Free
- Rating: 5.00/5 from 2 App Store ratings · 2 written reviews indexed
- Age rating: 4+
- Requires: iOS 17.4 · 21 MB
- Languages: American English
- Released: 2026-08-19
- Data updated: 2026-09-30
- Monetization: free
- User reviews in markdown: https://appshunter.io/ios/app/exponential-vibecode-together/id6788189402/reviews.md

## What is Exponential?

Vibecode together.

Exponential is a realtime issue tracker for teams that build with AI. Issues, customer support and coding agents in one workspace, in sync on every device.

Track work with your team:
• Issues with priorities, labels, due dates and custom statuses
• Boards for every part of your product, each optionally backed by a GitHub repository
• Markdown descriptions and comments with @mentions, task lists and inline images
• Realtime sync, the moment anything changes. No refresh, no stale views
• Push notifications for mentions, assignments and PR activity

Bring your own agents. Claude Code, Codex or pi run locally on your connected desktop, on your subscription. Pick an issue and hit Start coding, even from your phone. Your agent works on a real branch in your desktop IDE and opens the pull request when it is done.

• Watch coding sessions live in the Agents tab and steer them by message
• Answer your agent's questions right from your phone
• Save actions for repeating AI tasks like code reviews, deploys or merge conflicts, and run them with one tap

Customer support in the same workspace. Tickets from your users land next to your issues, replies go out by email, and the whole team sees the conversation.

Free for teams of up to three on the hosted cloud, one plan when you grow. Exponential is open source under Apache-2.0, so self-hosting is free at any company size.

## Key features

- Real-time issue tracking with priorities and custom statuses
- GitHub-backed boards for product development
- Markdown descriptions and comments with @mentions
- AI coding agents (Claude Code, Codex, pi) integration
- Live coding session monitoring and steering
- Integrated customer support ticketing
- Open-source and self-hostable

## Recent user reviews (2 of 2)

All indexed reviews: https://appshunter.io/ios/app/exponential-vibecode-together/id6788189402/reviews.md

### 5/5 — Awesome app

*2026-09-07*

As the title says… this app is awesome.

### 5/5 — Great for team coordination

*2026-09-05*

Excellent app for tech teams collaborating on projects, can connect to GitHub and is overall great for organization

## Frequently asked questions about Exponential

### What is Exponential's primary function?

Exponential is a real-time issue tracker designed for teams building with AI. It combines issue tracking, customer support, and AI coding agents into a single workspace, ensuring all devices stay synchronized.

### How does Exponential handle AI coding agents?

You can connect AI agents like Claude Code, Codex, or pi to Exponential. These agents can work on code directly from your connected desktop IDE, open pull requests, and their sessions can be monitored and steered via messages from your phone.

### Can I use Exponential for customer support?

Yes, Exponential integrates customer support tickets directly into the same workspace as your development issues. Team members can see conversations, and replies are sent via email.

### Is Exponential free to use?

Exponential is free for teams of up to three on the hosted cloud. It is also open source under Apache-2.0, allowing for free self-hosting at any company size.

### What devices does Exponential support?

Exponential is available on iPhone and iPad, ensuring you can track work and manage agents from multiple devices.

### How often is Exponential updated?

The latest version, 0.14.28, was released on September 10, 2026. This indicates active development and regular updates for the application.

### What is the age rating for Exponential?

Exponential has an age rating of 4+, making it suitable for a wide range of users.

### Does Exponential have ads?

No, Exponential does not contain advertisements. The application is free and does not rely on ad revenue for monetization.

## Version history (last 5 releases)

### 0.14.45 — 2026-09-26

- Reviews: every pull request row carries one control, Merge or Merge stack at the bottom of a stack, and a workflow node's pull request says it merges through the workflow
- Workflows: Open final PR appears when a workflow's final pull request was closed, the event log sits under the graph and a picked node shows only its own events, and failed or waiting states read amber
- Agent runs: a run waiting on your answer wears a red dot, and a workflow run on another account says which one
- Starting an issue that already has a live run says so instead of starting a second one

### 0.14.37 — 2026-09-19

- Device icons: give each machine its own icon in its settings, a monitor, server or laptop or the Apple, Windows or Linux mark, and pick from 96 icons for boards and actions
- The Run face leads with the agent's task list, a background lane keeps its tab while it runs, and the tabs and strips line up with the transcript
- Devices fold to one row each, show when a usage window resets, and a login can be removed from a machine; the Work screen's menu sits on the Issue face only, the face switcher counts the run's changed files and tucks into the expanded composer, and a batch run's issues open in a sheet
- Calls to Exponential read as their own group, Read 3 issues instead of a generic tool run, and a turn that changed no file no longer shows an empty edits card
- The Stop button on a running session is a plain bar button now, the same height as the back button, instead of a pill that drew a broken outline
- Rows in a run's transcript no longer overlap each other, and the chat scrolls clear of the card above the composer
- A chat run takes its name from what the agent is working on, instead of reading Chat everywhere
- One diff everywhere: a diff's files sit in a tree you can jump around, and an agent's edits in a turn collect in one Edited files card, with line numbers, hidden unchanged lines you can open, and renames and binary files named
- Usage in one place: an overlay shows the run's own account, its limit windows and the context window, plus the other logins on that machine
- Devices carry the logins: every machine lists its agent logins with plan, health and usage, and a device's gear button holds Update and Remove
- One issue search: the same ranking drives every issue list, picker and # menu, with the top result ready to pick
- A running command shows only its last lines, and a run that goes quiet is marked ended and stays in the list
- One Work screen: an issue, its coding run and its changes are faces of one screen, switched from the bar at the bottom, with Stop or Resume in the top-right corner
- Stacked pull requests: start an issue on top of the one that blocks it, merge the chain with Merge stack, read the stack or batch badge on the work, and answer a question a nested run sends you
- Results from a run: a run that publishes screenshots gains a Results face, grouped by topic, one tile per platform
- Batch runs by name: a run covering several issues reads EXP-874 +2 beside the first issue's title instead of Batch run
- Messages that land at once: a message typed while the agent works reaches it immediately and shows as sent until it is read, a waiting message shows in a bar above the composer, and Stop hands the text back to your composer
- GitHub, sorted out: connecting GitHub, adding a repository and setting a board's repository read and behave the same as on the desktop, with errors shown where you are
- Quieter buttons and clearer pickers: only a screen's main action keeps a filled circle, and the agent picker carries the Claude Code or Codex mark while device menus lead with a laptop or server icon
- The Agent page holds your runs: past runs fold behind a count you can open, and with nothing running the prompt sits in the middle of the page
- Issue filters are gone: search covers what they did, and grouping and the status bands stay
- Continue with email, or with a passkey: sign in with a 6-digit code mailed to you, or with Face ID or Touch ID; every option is one Continue list
- Switch account mid-run, and come back where you were: a Claude run at its limit can carry on with another account, and leaving a session returns you to the issue or board you opened it from

### 0.14.30 — 2026-09-12

- Videos and audio in descriptions and comments: attach a clip and it plays inline, with a poster frame and its length; what you upload is converted to 720p first
- GitHub accounts: team settings lists every connected account with a Configure link, and Connect another account opens GitHub's own picker, so a second organization can be installed
- Add repository: the picker says why a repository is missing, links to each account's GitHub settings, and takes a repository by name
- Shorter past runs: an ended run reads the machine and the time, without the agent name and the reason it ended
- The running session dot moved to the Chat button in the bottom bar, which now shows on every tab; Devices keeps only the sign-in warning
- The composer always names the machine a run will start on, and opens a menu only when there is more than one
- Rate limit banners clear themselves once the reset time has passed, and a plain usage warning never shows as one
- Share a machine with several teams: the device sheet has one switch per team instead of a single choice
- Composer details: Plan is a switch, a picked issue or action is removed with its own close button, and a board's bottom bar merges chat and new issue into one capsule
- Pins: pin an issue, a session or an action from its menu, and it sits in a Pinned section at the top of the board switcher
- Accounts on the Devices page: one row per agent account with its usage bars and a chip per machine, the same section web and desktop have
- One composer: the Agent page composer starts everything, with issues or an action picked at the top and the machine, agent, model and plan mode under it; the Start coding sheet is gone
- Every play button lands there: Start coding on an issue, the bulk bar, an action's Run and Fix merge conflicts open the composer with the right things already picked
- Instructions with a start: whatever you type beside the picked issues or action reaches the run as extra instructions, and for a chat it is the prompt
- Images on a start: attach up to four images to the composer and the machine has them before the first turn
- Actions keep only picks: an action asks for a repository, board, pull request or icon; the free text comes from the composer, with the action's own hint as the placeholder
- Sessions beside the composer: the Agent page lists your running and past runs next to the prompt, and a run started by another run is nested under it
- Answer inside the card: a plan or a question is answered in its own card, with an inline field when you want to write instead of pick; the composer steps aside until the card is done
- Agent messages: an agent can send your team a notification over MCP, and it arrives as a push and its own row in the Inbox
- Rate limits that add up: a usage warning no longer marks a run rate limited, and a blocked run's window and reset time now describe the same limit

### 0.14.28 — 2026-09-10

- Rate limited, not stuck: when an agent runs out of usage mid-run, the session is marked as rate limited with the time it resets, and stays live and steerable instead of going quiet
- Steer composer: the session field takes @ for teammates, # for issues and : for emoji, the same three the comment box already had
- Edits inline: an edit's diff renders under its row in the transcript, folded away until you tap it
- Tool groups that say what happened: a collapsed run of tool calls reads "Ran 4 commands, edited 2 files, 1 failed" instead of counting calls
- A start that would go nowhere is refused: starting on a machine whose agent is already out of usage says so and names when it resets
- Usage bars for a second agent account on the same machine
- Steering: switch between Plan and Build from the live session; model and effort are chosen once when you start, and context and usage readouts update as the run reports them
- Past runs: finished sessions are listed per machine with Resume, so an ended run can be picked up again
- Chat without a repository: a chat started from the Start coding sheet no longer needs a repository and runs in a scratch folder with only the Exponential tools
- Agent picker: when an agent is not ready on a machine, the picker says so and blocks the start until you run the doctor there
- Steering feed: issue identifiers like #EXP-12 become tappable chips, and agent commands render as command pills
- Issue relations: link issues as parent, sub-issue, blocking, blocked by, duplicate or related from the properties sheet; typing #EXP-12 in a description or comment links the two issues on its own
- Threaded comments: reply under a comment from the docked composer; threads stay one level deep, and a comment an agent posted over MCP says so in its header
- A guided first run: creating a team walks through the first board, inviting teammates and setting up devices, in the same order as web and desktop
- Team settings: owners can create invite links from the Members section
- Merge from action and chat runs: a run that opened a pull request shows Merge in its steering view, and Reviews lists those pull requests under Agent runs
- Bigger comment images: pictures attached to a comment show at full width and open the viewer on tap
- A quieter activity feed: events read as one muted line with a time, comments sit in a single card
- Subscribe button retired: you follow an issue automatically when you create it, comment, get assigned or are mentioned
- Steering: the context bar no longer overlaps the transcript
- Sessions stay on your machine: transcripts are stored on the device that ran them, never on the server; past runs open from Devices, Chat and Automations and are fetched from that machine when it is online
- Subagent conversations: a subagent's turns show in their own view, and agent prose no longer splits mid-sentence into separate rows
- A run that goes quiet says so: the session shows when the agent has stopped reporting progress
- Remote agent sign-in: paste the code from an agent's login link into the machine's settings to finish signing it in from your phone
- Start coding: one sheet for issues, actions and chat with the same options everywhere

### 0.14.24 — 2026-09-04

- Tables: write tables in issues and comments, with column alignment; edit cells on your phone and long-press to delete a whole table
- Steering: type "/" to run the agent's built-in commands such as /compact and /clear, and watch the agent compact its context instead of guessing
- Comments show @mentions as member pills instead of the raw address
- Boards: every board can set its own default branch in board settings; coding sessions start from it
- Review diffs scroll sideways with a compact header per file
- Empty boards show an icon, a title and a hint
- Settings, action and start-coding sheets follow the platform idioms more closely
- A refreshed glass look across lists, sheets and controls
- Todo and Backlog are now one list: everything that sat in Todo lives in Backlog
- The editor keeps your intentional blank lines, identically on every platform
- Agents always run without permission prompts; the toggle is gone
- One agent picker: agent, model, effort, plan mode, account and usage in a single card
- Actions: create and edit actions with their inputs right on your phone
- Reviews: a new list and PR diff with per-file cards; Fix merge conflicts replaces Merge when a merge hits a conflict
- Navigation badges are dots: Inbox and Support on unread, Devices while a run is live, Reviews while a PR is open
- Steering: plans are no longer cut off or shown twice, and a finished session closes its view
- Starting a run on an offline machine is refused right away instead of being silently queued

## Apps similar to Exponential

| App | Rating | Price | Category |
| --- | --- | --- | --- |
| [Macro App](https://appshunter.io/ios/app/macro-app/id6743133649) | 3.8 (8) | Free | Productivity |
| [Nimbalyst](https://appshunter.io/ios/app/nimbalyst/id6756393105) | 5.0 (5) | Free | Productivity |
| [Markdown Editor & Reader](https://appshunter.io/ios/app/markdown-editor-and-reader/id6768164089) | — | Free | Productivity |
| [u-Parking](https://appshunter.io/ios/app/u-parking/id6446409248) | — | Free | Productivity |
| [Unzip Fast - ZIP RAR 7Z](https://appshunter.io/ios/app/unzip-fast-zip-rar-7z/id6752353634) | 1.0 (1) | $0.29 | Productivity |
| [SpeedTest: WIFI Speed Test](https://appshunter.io/ios/app/speedtest-wifi-speed-test/id6741730996) | 4.3 (13) | Free | Productivity |

---

*Data collected daily from the US App Store and indexed by [AppsHunter](https://appshunter.io/). User reviews are verbatim App Store reviews. Ratings, prices and chart positions refresh continuously; this snapshot is from 2026-09-30.*