> ## Documentation Index > Fetch the complete documentation index at: https://docs.tasksmind.com/llms.txt > Use this file to discover all available pages before exploring further. # Your setup guide > Everything you and your team do to turn customer issues into fixes, automatically. When one of your customers reports a problem (or your team spots one), tasksmind turns it into an incident, investigates the root cause in your code, writes a fix, and opens a pull request (a proposed code change) on your repository. Your team is kept in the loop in Slack the whole way, and one of your engineers approves the change before anything ships. tasksmind never merges on its own, so you stay in control. The steps below are in the order you'll complete them, grouped into three short phases. Most take a couple of minutes. A couple of steps, like getting your account activated, we handle together with you during onboarding, so you're never stuck. Your tasksmind contact is with you the whole way. ## Your setup at a glance Eight steps across three phases. Each one is detailed later in this guide. **Phase 1: Your account & team** | # | Step | What it's for | Who does it | | - | ------------------------ | --------------------------------- | ----------- | | 1 | Create your account | Sign up + confirm email | You | | 2 | Set up your organization | Create or join your company's org | You | | 3 | Invite your team | Add teammates with roles | You | **Phase 2: Get connected** | # | Step | What it's for | Who does it | | - | --------------------------------- | --------------------------------- | ----------------------- | | 4 | Get your account activated | We set up your access with you | With the tasksmind team | | 5 | Install the GitHub App | Grant access to your repositories | You | | 6 | Connect your repo & add a project | Point tasksmind at the repo | You | | 7 | Connect Slack | Your customers' front door | You | **Phase 3: Go live** | # | Step | What it's for | Who does it | | - | ---------------------- | ------------------------------------- | ----------- | | 8 | See it work end to end | A real issue → a fix, in front of you | You | ## Phase 1: Your account & team Who does this: the first person from your company (you'll become the org owner). 1. Go to the sign-up page and choose **Create account** (or continue with Google). 2. Enter your name, work email, and a password (needs an uppercase letter, a number, and a symbol). 3. Complete the quick "prove you're human" check. 4. Open the confirmation email and click the activation link. You'll need to confirm before using the app. Then: you're taken straight into a short setup wizard (Step 2). No invite or beta code needed. Why: tasksmind is built around your organization. Your access, repositories, and integrations all live there, shared by your whole team. So the first thing you do is create your company's organization (or join one a colleague already made). Choose **Create a new organization.** Enter your organization name and your company email domain (e.g. `acme.com`). Optionally invite teammates now, or do it later in Step 3, then finish. Choose **Join an existing organization.** Paste an invite link, or enter your company email so an admin can approve you. Who can do this: owners and admins. Full control of the organization. The person who created it. Can invite and manage members and settings. Can't remove the owner. Uses the product day to day. Open your Organization page and go to member management. Enter teammate emails to send invites (each link works for 7 days). Teammates click the link to join, or request to join with their company email and you approve them. ## Phase 2: Get connected This phase gives tasksmind what it needs to work for you: an activated account, access to your repository, and a place to hear from your customers. What this means: your usage and access are set up together with the tasksmind team. Reach out to your tasksmind contact to get your account activated. You can go ahead and set up the tools below in the meantime. **We'll handle this with you.** During onboarding, your tasksmind contact activates your account so everything is ready when your first issue comes in. Nothing here is self-checkout, just a quick conversation. Why it's first: the tasksmind GitHub App is how we securely read your code and open fix pull requests. Installing it is also where you grant access to the specific repositories tasksmind should work on. 1. In **Settings → Connect Your Tools**, click **Install App** under GitHub App. 2. On GitHub, pick the account or organization to install on, and choose which repositories tasksmind may access. 3. Approve. GitHub sends you back to tasksmind and the connection completes automatically. Fixes always arrive as pull requests on branches named `tasksmind/…`, proposed changes your team reviews and merges. tasksmind never merges by itself. Once the GitHub App is installed, connect the repository you granted access to and create a project. A project simply points tasksmind at one repository and holds your preferences for it. 1. Connect the repository you just granted access to. 2. Create a project: give it a name, confirm the repository and the branch to base fixes on (usually `main`). 3. Set your preferences. Your team's approval is always required before anything merges. **New to this? We'll get you started.** Your tasksmind contact can set up your first repository and project with you during onboarding, so your very first issue has somewhere to go. Why Slack matters most: Slack is your customers' front door. Connect it so the questions your customers raise in a Slack channel become incidents automatically. tasksmind investigates each one and, when the fix ships, replies to them with a friendly "this is fixed" note. 1. In **Settings → Slack integration**, click **Add to Slack** (owner/admin) and approve the permissions. 2. Invite the agent into your customer support channel: `/invite @tasksmind` 3. Connect that channel to your repository: `@tasksmind connect support acme/web` (owner/admin). Customers here only ever see friendly updates. 4. Keep your engineers posted: in your team's internal engineering channel, run `@tasksmind set dev channel here`. Root cause, file paths, and PR links go here, never to the customer channel. (The agent will refuse if you run it in a support channel.) Just @mention the agent. That's how you give it every instruction. See the [Slack agent reference](/reference/slack-agent) for the full command list. ## Phase 3: Go live With your repository and Slack connected, watch a real issue flow all the way through: 1. A customer posts a problem in your connected support channel. 2. tasksmind turns it into an incident and starts investigating. Your team sees progress in your internal channel. 3. It pinpoints the root cause, drafts a fix, and opens a pull request. 4. One of your engineers reviews the pull request, approves it, and merges. 5. tasksmind replies to the customer in their original thread to let them know it's been fixed. **You're live.** A real customer issue now flows in, becomes a pull request your team can merge, and the customer is notified automatically. The only manual step is reviewing and merging the change. ## A closer look: what happens inside A plain-language tour of the stages every incident moves through, for anyone who's curious. 1. **Incident created.** A customer's message (or a report from your team) creates the incident and posts an "investigating" update for your engineers in Slack. 2. **Investigation.** tasksmind reads the report and your code to pinpoint the root cause. If the report is too vague, it asks a clarifying question in the customer's thread rather than guessing. 3. **Fix.** It drafts the change and opens a pull request. Riskier changes are flagged so they're never merged without a careful look. 4. **Your approval.** One of your engineers approves and merges. tasksmind never merges on its own; that stays with your team. 5. **Merge & deploy.** Once merged, the incident is marked fixed. If you connect a deploy tool, tasksmind waits until the fix actually ships to production before telling the customer. 6. **Safety nets.** If a fix gets stuck, tasksmind retries automatically, and anything it can't handle safely is flagged for a person. A customer's reply to a question picks the investigation right back up. ### Statuses you may see on an incident | Status | What it means | | ------------- | ------------------------------------------------------------------------------ | | Investigating | Being looked at right now. | | Investigated | Root cause found; ready for a fix to be drafted. | | Awaiting info | tasksmind asked the customer a clarifying question and is waiting for a reply. | | Fix ready | A pull request is open and waiting for your approval. | | Fixed | The fix was merged (and deployed, if you've connected a deploy tool). | | Needs human | tasksmind couldn't proceed safely and has flagged it for a person. | ## Questions at any point? Your tasksmind contact is here to help you through every step. Include the step you're on and any error text you're seeing. Real walkthroughs of the product in action.