Do you often clock in to many different little tasks? Are you annoyed that you can’t just clock in to one of your most recent tasks after restarting Emacs?

The function org-mru-clock-in from this package will pre-fill your clock history with clocks from your agenda files (and other open org files) so you can clock in to your most recent clocks regardless of whether you just started Emacs or have had it running for decades. Tasks are sorted by recency, and uses completing-read for quick selection. This makes it a nice replacement for org-clock-in-last.

It uses completing-read-function (overridable with org-mru-clock-completing-read) on org-mru-clock-in to make clocking in faster. If you set that to ivy-completing-read, you can hit M-o g to visit to the task heading instead of clocking in.

The list is sorted to keep the most recently clocked entry first, although if the entry at point is an org-mode heading (in an org file or the agenda), then that will be pushed to the top of the list (you can turn off this behaviour by setting org-mru-clock-include-entry-at-point to nil).

You can also use org-mru-clock-select-recent-task as a replacement for org-clock-select-task, again with pre-filled history.

You may also capture new tasks on the fly if your search text didn’t match anything – try C-h v org-mru-clock-capture-if-no-match RET.

If you use MELPA, you can just do M-x list-packages, find org-mru-clock in the list and hit i x.

Just put org-mru-clock.el somewhere in load-path.

To use, require and bind whatever keys you prefer to the interactive functions:

(require 'org-mru-clock)
(global-set-key (kbd "C-c C-x i") #'org-mru-clock-in)
(global-set-key (kbd "C-c C-x C-j") #'org-mru-clock-select-recent-task)

Maybe trade some initial slowness for more tasks cached:

(setq org-mru-clock-how-many 100)

But don’t set it higher than the actual number of tasks; then it’ll always try (and fail) to fill up the history cache!

If you want to use ivy for org-mru-clock-in:

(setq org-mru-clock-completing-read #'ivy-completing-read)

If you prefer use-package, the above settings would be:

(use-package org-mru-clock
  :ensure t
  :bind* (("C-c C-x i" . org-mru-clock-in)
          ("C-c C-x C-j" . org-mru-clock-select-recent-task))
  :init
  (setq org-mru-clock-how-many 100
        org-mru-clock-completing-read #'ivy-completing-read))

If you do

(setq org-mru-clock-keep-formatting t)

then entries will be shown in their org-mode faces, instead of whatever face your org-mru-clock-completing-read function uses by default.

If you do

(setq org-mru-clock-include-entry-at-point nil)

then the entry at point will not be prepended to the start of the list.

By default, all open org-mode files are searched for recent clocks, but you can change this to e.g. only the org-agenda-files with

(setq org-mru-clock-files #'org-agenda-files)

or some other function of your own devising.

You can also exclude clocks by setting org-mru-clock-predicate to e.g. org-entry-is-todo-p (which will exclude DONE tasks) or org-mru-clock-exclude-done-and-archived (which will also exclude those tagged with org-archive-tag).

You may also be interested in these general org-clock settings (documented in the Org-mode manual):

(setq org-clock-persist t)
(org-clock-persistence-insinuate)