TurtleTask

Documentation

Everything you need to get a team planning in TurtleTask. New here? Start at the top; the whole thing takes about five minutes to read.

Getting started

  1. Create your organisation. Head to app.turtletask.app and sign up. The first account becomes the organisation's admin.
  2. Add your first task. Open the Tasks view and use the add button — you drop straight into editing. Give it a team, project, assignee, a duration in working days, and a start date.
  3. Invite your team. From Settings → Users, add people by email, or connect single sign-on so they log in with your company accounts.
  4. Watch it schedule itself. Add dependencies between tasks and TurtleTask works out the dates for you.

Core concepts

A task carries a team, project, assignee, stakeholder, status, priority, weight, a duration in working days, a start date, and any dependencies.

  • Working days. Durations and schedules are counted Monday–Friday. Weekends never stretch a timeline, and the numbers match the spreadsheet formulas your team is used to.
  • Statuses. Not Started, In Progress, Blocked, Done, and Archived. Archived tasks are hidden everywhere by default; one toggle reveals them.
  • Weight (1–5). A task's share of a person's day. Someone splitting time across several tasks has their capacity divided by weight, so allocation never double-counts.
  • Out of office. Time off is modelled directly — it shows as gaps in schedules and absorbs capacity, with no fake "vacation tasks" cluttering your plan.

The views

Every view is a lens on the same live plan, and they all share one filter bar — multi-select chips for team, project, assignee, and stakeholder, plus week / month / quarter / year presets. Press / anywhere to search, and g then a letter to jump between views; ? lists every shortcut. On a phone, the same plan gets a card list, a swipeable board and a full-screen task drawer.

  • Home — where the app opens. Lists of work that should have started, work running past its end date, requests waiting in the inbox, and a line per person showing what they are on and what is late. Each tile acts: advance a status, defer by working days, size unscheduled work, unblock, archive — all undoable.
  • Tasks — a fast, editable grid with keyboard-friendly autocomplete and colour-coded statuses. Click any task ID to highlight everything it depends on and everything waiting on it.
  • Kanban — drag work between statuses. If a card's assignee is away today, the card says so.
  • Map — the plan as a dependency map: tasks are cards, dependencies are arrows. Drag from one task to another to connect them, select a task and set its status from the toolbar, delete an arrow to remove a dependency, and add follow-on work with one click.
  • Gantt — timelines on a working-day axis, with today marked and time off shown as gaps. You choose the dimensions: rows by assignee, project, stakeholder, or team, and bar colours by any other of those. Drag a bar to move it or its edge to resize it, and everything downstream re-flows.
  • Bars — weighted person-days as stacked bars. Group and stack by any combination of assignee, project, stakeholder, or team to see who is over- or under-committed.
  • Reports — how the plan moved, rebuilt from its own history: each project's projected finish over time, scope against done, flow by status, and weekly throughput. A per-task ranking names what moved the date, including tasks pushed by something upstream.

Scheduling & dependencies

A task can depend on several others; its start snaps to the working day after its latest parent finishes. Change one task's dates or duration and every task downstream reschedules instantly — including long chains and diamond-shaped dependency graphs, where a task correctly waits for its latest parent.

Dependencies can be typed into the grid or drawn on the Map — and when you connect two tasks there, the new downstream task even inherits blank team and project fields from its parent. There are no formulas to maintain and no dates to chase by hand: you describe the dependencies once, and TurtleTask keeps the schedule correct as things move.

Time travel

Every change to every task is versioned. Use the As of control to pick any past date and time, and the entire app — grid, kanban, gantts, allocation — shows the plan exactly as it stood then, including tasks deleted since. A banner marks snapshot mode, which is read-only.

Combine it with the date-range filter to ask questions like "as of June 1, what was scheduled for July?" — and put the schedule before and after a slip side by side.

Team & access

  • Roles. Admins manage users, API keys, single sign-on, and organisation settings. Normal members get full use of the tracker. Role changes take effect immediately.
  • Membership is admin-only. There are no self-serve invite links: people join because an admin created their account under Settings → Users, or because they signed in through your identity provider from one of the email domains you registered.
  • Single sign-on. Connect any OIDC provider — Google, Microsoft Entra, Okta, Auth0, Keycloak. New users are provisioned automatically on first login; deactivating someone cuts their access everywhere, at once.
  • Two-factor sign-in. Anyone signing in with a password can add an authenticator app, and admins can require it for the whole organisation. People who sign in through your identity provider follow its policy instead.
  • API keys. Named, revocable, optionally read-only, limited to matching projects or teams, and given an expiry. Every change a key makes is attributed to it by name.
  • Isolation. Your organisation's data is completely separate from every other organisation's.
  • Everything is yours to take. Settings → Data exports the whole organisation — tasks, comments, attachments, people, calendar and the full change history — as one zip you can import elsewhere.

Configure all of this under Settings once you're signed in as an admin.

Security

  • Encryption. Traffic is encrypted in transit and data at rest. Passwords are hashed; API keys and session tokens are stored only as hashes.
  • Isolation. Organisations are fully separate from one another — tasks, people, attachments, history and API keys are all scoped to yours.
  • Backups. Taken nightly and stored away from the application host.
  • Billing. You pay per active user, and only admins add people, so the count changes only when you change it. Admins are emailed when someone is added or removed, and if you'd like a hard cap on seats, ask and we'll set one.
  • Your data. It's yours. Export the whole organisation at any time from Settings → Data; ask us to delete it and we do. We don't sell it or use it to train anything.
  • Found a vulnerability? Email [email protected] with "Security" in the subject line — also listed at /.well-known/security.txt.

API & integrations

Everything the interface does is available over a REST API: tasks, bulk updates, reordering, comments, history, charts, and time-travel queries. Admins can mint named, revocable API keys (Settings → API Keys) for scripts and integrations, and every change is attributed in the audit trail to the person or key that made it.

For quick, one-off needs there's also a what-you-see CSV export in the menu — it respects your current filters (and even the "As of" snapshot), so the file matches exactly what's on screen.

Interactive API reference (Swagger and ReDoc) is built into the app — open it here once you're signed in.

Working with AI agents? Your workspace serves a ready-made Claude skill at /api/skill.zip — drop it into Claude Code or claude.ai and Claude knows how to drive your board: list and create tasks, update schedules, comment, and pull history for progress reports. Give each agent its own API key so its changes are attributed like anyone else's.

Ready to try it?

Get started