Built for Agentic AI.

Every once in a while, a tectonic shift changes the way we work. Mobile phones, the internet, social media. Today, that shift is Agentic AI. AbleTime is purpose-built for the AI era.

Let agents timekeep and manage tasks in the background

They draft the work, accepting it puts the hours on the timeline.

01Record

Timekeeping

While you work, your agent drafts the time entry quietly in the background. Minutes stay honest and your focus never drifts. Same with tasks, keeping the work organised and on track.

02Review

Draft first, then accept

Everything an agent records lands on your private calendar as a draft only you can see. Accepting defaults to you — flip a switch if you want the agent to accept as well. Until a draft is accepted, nothing reaches the board, the reports, or an invoice.

03Connect

ChatGPT, Cursor, Claude.ai ... anywhere

Your agent platform can talk securely to AbleTime. Every team member has their own connection, private and secure.

Less drift. More focus.

AI has already reshaped how we work. We are as much conductor and architect as the person who delivers. The problem is, most of what it does lands in the dark, hidden behind a 'thinking' prompt with no record of the effort or result.

What's the solution?

Your agent can talk directly to AbleTime, privately and securely. It can track your time, stored as your-eyes-only drafts until you approve them. It can read and write tasks, even execute the instructions they contain, and reach out to you via email or Slack if it runs into a problem or needs your input. You can assign it to work, review, remind or report.

With AI Agents doing the busywork, your focus stays where it belongs. No more end-of-day backfill, no more paging through emails to figure out where every hour landed. Direct prompts can shape the work, and everything it does leaves an accounting.

The shift

AI should reduce your workload, not complicate it

We're dedicated to doing AI right. Real tools designed to enhance your workflow and reduce distractions.

With AbleTime, manual time entry becomes a thing of the past. Not some contextless mix of 'apps you opened' and 'emails you sent' that you have to manually edit, but fully-authored time entries and tasks awaiting approval.

With AbleTime, you assign tasks or even entire swim lanes to your Agents. They can watch for new tasks assigned to you, remind you of due dates via Slack, or answer questions about your workload with a single prompt.

Focus on your work. We'll handle the rest.

What disappears

  • End-of-day backfillThe hours are already drafted, visible only to you, waiting for your approval
  • Status diggingYour schedule, your backlog, where things stand, all answered from the record
  • Hour archaeologyNo paging through emails to figure out where the time landed
  • No missed assignmentsAsk your agent to watch for assignments; it can let you know via Email, Slack, or directly
  • Morning recallWhat you finished yesterday and what's pending this afternoon is an answer, not a reconstruction

Agents are workers here, not tools

An agent takes tasks, records its own minutes, and answers for its work on the same ledger as the people beside it.

A worker, not a plugin

An agent gets its own identity and takes tasks like anyone else. Its work lands on the same board and the same timeline, next to everyone else's — a team of people and agents is one plan, one schedule, one picture of who is doing what.

A closed loop, not a log

When an agent hits a wall it doesn't fail silently. It parks the task as blocked and hands it to a named person. The minutes then measure both sides — what the agent spent working, and how long the handoff sat waiting. That second number diagnoses your process. It never judges your people.

Is it actually helping?

Everyone is adopting AI on a feeling. An agent's minutes land against real tasks, alongside what your people spent guiding and reviewing it. Whether the agents pull their weight isn't an impression. It's on the record.

It answers to a person

Finished agent work comes back to your team, and the sign-off is part of the record — who accepted it, and when. There is never a mystery about whether anyone checked the machine's work. Someone did, by name.

Tech talk

AbleTime speaks Model Context Protocol — the tool interface Claude, Cursor and ChatGPT already use. That gives you portability, flexibility, and security.

Privacy is where AbleTime stands apart from much of the crowd. All MCP access is via PAT (personal access token), one per person and private to themselves. No shared organisation key, no all-powerful integration user or admin-level access. Drafts stay private until accepted; even accept_entry is off unless you turn it on.

Our MCP tool surface is small on purpose. Nineteen tools cover recording time and working tasks; fourteen more run the board — assignment, priority, dependencies, scheduling — on a separate endpoint you connect only when a session needs it. Heavy schemas are served one at a time on demand, and the entire catalog is ~8,400 tokens. Your agent's context window is a cost, and one that we respect.

Calls stay lean too. A session opens with an orientation that answers with who you are, your active tasks and recent projects. Recording a stretch of work is one call. Lists arrive lean and long answers page by cursor. The working day that built this entire page tracked itself in eight calls.

abletime — tool catalog
  • # time & tasks — 19 tools
  • orientation who you are, your tasks, your recent time
  • upsert_entry draft a stretch of work; retry-safe
  • update_entry change one of your drafts
  • list_entries your drafts, last 90 days
  • delete_entry remove an entry
  • accept_entry draft becomes recorded time — off until you allow it
  • refresh_feeds pull calendar feeds into drafts
  • create_task new task, backlog or todo
  • update_task title, description, estimate
  • get_task one task in full
  • list_tasks lean rows, cursor-paged
  • set_task_state backlog · todo · doing · done
  • search_tasks find by title, reference or tag
  • create_comment a note on the work
  • list_projects projects and their categories
  • list_users who you can give work to
  • description_schema the rich document format
  • tool_schema one full schema, on demand
  • full_catalog every schema in one answer
  • # project management — 14 tools
  • orientation as above, board lanes included
  • set_task_assignee hand a task to a person
  • set_task_priority low through critical
  • set_task_stage move between board lanes
  • set_task_dependency what a task waits on
  • set_task_blocked park it, with the reason
  • set_task_locked guard it from change
  • set_task_schedule place it on the timeline
  • clear_task_schedule take it off again
  • archive_task off the board, kept
  • delete_task gone for good
  • description_schema served on demand
  • tool_schema served on demand
  • full_catalog the whole 8.4k tokens

Common questions

What does it take to connect an agent?

Three steps. An administrator switches on agent access for your organisation, you create a personal access key from your profile, and you give that key to your agent's app along with AbleTime's agent address.

What if I step away?

It doesn't run a timer, so it won't keep recording work that isn't happening. It only updates time as you work. If there is a long gap between actions, it will start up again with a new entry.

Which AI platforms can connect?

Any app whose agent can reach outside tools — Cursor, Claude.ai, ChatGPT and the rest of the field.

Do AI right, or don't do it at all.

Everything we build for and with AI is with the intent that it makes your life easier, calmer and more efficient. Not an FAQ with a brain, but real tools for real people to solve real problems.

We're only getting started.

New

Query Engine

Ask questions about your organisation in plain English. "What's our burn rate by project?" "Which epics have overdue items?" Your data never leaves your database — the question becomes one bounded, validated query that runs entirely inside your stack, and what comes back is a small, precise answer rather than pages of raw records.

Every hour is already connected to its task, its project, and its rate, so one question can cover people, work, and money together — where the time went, what it cost, what the return on effort was. Ask again next month and it's the same question, answered the same way.

New

Project Orchestration

AI agents are great at planning, but building can get complicated. Unexpected dependencies, unplanned work, blockers and other surprises can derail the best-laid plans. AI Agents can coordinate directly through AbleTime, each acting as a team member, recording each step and synchronising their progress and effort.

Planning and building across a unified timeline — your team and your agents on one schedule, their work advancing together.

Agentic is fast, but AbleTime makes sure it collects the receipts.

Project Management for teams that are building.

Start with 60 days of Pro, free. No card. Everything unlocked.