Help center
How CodeLynq works and how to use it: projects, tasks and specs, runs and pull requests, the architecture map, deployments, billing and security.
Getting started
- What CodeLynq is
How a request becomes a reviewed pull request - the whole flow from task to spec, approval, run, pull request and deployment.
- Getting started
From a new account to your first task - sign up, create an organization, choose a plan, create a project and connect a repository.
- Navigating the app
What is where - the sidebar, the dashboard, the inbox, switching organization or language, and where to find help.
- Voice input: dictate tasks instead of typing
Dictate a task, an answer or a description instead of typing it, in the app and on your phone, transcribed and ready to edit before you send it.
Projects and repositories
- Projects
Create a project, give CodeLynq the context it needs about your code, and let it fill that context in from your repository.
- Connect a repository
Connect a GitHub, GitLab or Bitbucket repository to a project, set its default branch, check the connection and set up the CI webhook.
- Start a new application
No code yet? A short scoping interview turns your idea into architecture decisions and a build plan that CodeLynq builds in an empty repository.
- Maintained branches
Tell CodeLynq which branches of a repository are live, and whether new tasks apply to all of them by default.
- Codebase memory: what the AI agent remembers between runs
What CodeLynq remembers about your code between runs, how to add your own notes, and organization-wide standards every run follows.
- Architecture map and rules
See the structure of your code, zoom from components down to classes, check what each run did to it, and hold every run to the architecture you intend.
Tasks
- Tasks
Create tasks, choose their type and priority, report bugs well, link dependencies, archive tasks and follow their status.
- Create a task by chatting
Describe what you need in your own words - typed or spoken - and the assistant turns it into a clear task.
- Work on several branches at once
Apply one task to several version or release branches, or start a new version line - with one pull request per branch.
- Technical spec (TSD) review and approval
Review the technical solution design CodeLynq proposes, confirm the decisions it asks you about, put your answers into the spec, and approve it to start building.
- Mockups: see the screens before they are built
Let CodeLynq design the screens a spec describes, give feedback until they are right, and approve them with the spec so the run builds to match.
- Troubleshooting production issues with a read-only AI chat
Investigate a problem with a read-only debugging chat that can read your code, logs and database, and turn its diagnosis into a fix task.
Runs and pull requests
- Runs
Start a development run, follow its progress and logs, stop or retry it, and find the pull requests it opened.
- Quality gates for AI code: tests, review and security scan
The checks a run goes through before it opens a pull request - your verify command, a self-review, a security scan and documentation updates - and which apply on self-hosted runners.
- The CI feedback loop
Let CodeLynq address failing CI checks and review comments on its pull request - on request, or automatically until the checks are green.
- Approving expensive runs
How CodeLynq estimates what a run will cost, and how spend limits make expensive runs wait for an admin's approval.
Planning
- Plan a larger goal as an epic
Describe a goal, let CodeLynq break it into tasks with dependencies, approve them, and run them in the right order.
- Turn meeting notes into tasks
Paste notes from a meeting, or import them from Fireflies, and let CodeLynq pull the work out of them as tasks.
- Groom the backlog with AI
Let CodeLynq propose priority, effort and order for a project's open tasks, then review and apply the changes.
- Inbox and dashboard
See at a glance what needs a decision across all your projects, and where your work stands.
- Analytics and costs
What your team delivered, what runs cost, and where to find it - per run, per project and for the whole organization.
Deploy and operate
- Deployments and releases
Connect your existing deployment pipeline, deploy from CodeLynq, and deploy automatically when a CodeLynq pull request is merged.
- Hosting your application
Let CodeLynq provision a database and run your application, on infrastructure it manages or on your own Scaleway account.
- Turn monitoring alerts into tasks
Point your monitoring tool at CodeLynq so production alerts become bug tasks - and, if you allow it, get a guarded fix automatically.
Organization and team
- Organizations and members
Create an organization, switch between organizations, invite your team and accept an invitation.
- Roles and permissions
What organization admins, members and project owners can do in CodeLynq: who approves specs, starts runs, changes settings and manages billing.
- Plans, credit and invoices
Choose a plan, keep model credit topped up, set spending limits per run and per month, approve expensive runs and find your invoices.
- Single sign-on (SSO), MFA and access security
Two-factor authentication and sessions for your own account, and single sign-on, SCIM, MFA enforcement, IP allowlists and audit logs for your organization.
Integrations and runners
- Integrations: GitHub, GitLab, Bitbucket, CI, Slack and MCP
Everything CodeLynq connects to: GitHub, GitLab, Bitbucket, CI, Slack, webhooks, meeting tools, monitoring, deployments and MCP, and where to set each up.
- Start tasks from Jira, Linear or Azure DevOps
Put a label on a ticket and CodeLynq takes it from there - the spec, the approval, the pull request and the status all show up on the ticket.
- API tokens
Create personal tokens for scripts and tools that use the CodeLynq API, and manage connected AI assistants.
- Connect an external AI assistant (MCP)
Let Claude or another MCP-capable assistant create tasks, approve specs, start runs and deploy - with the access you approve.
- Self-hosted runners: run the AI agent on your own machines
Run CodeLynq's development work on your own machines - register a runner, install it, run several jobs at once, serve several organizations and keep it up to date.
Mobile app
- The CodeLynq mobile app
Follow your projects, create tasks by chatting or speaking, approve specs and spend, and start runs from your phone.
Test it on a task from your own backlog
Send us one real task. We run CodeLynq on your repository and you review the pull request. Free, within 5 working days of approving the spec.