Running the script
For a short description of all available options, run
ephemetoot --help from the command line.
It is strongly recommended that you do a test run before using
ephemetoot live. There is no "undo"!
Create your config file (--init)
Before you can use
ephemetoot you need a configuration file. You can create this yourself, but
--init will walk you through the values you need and save it to
config.yaml in the directory from which you run the command:
Run in test mode (--test)
To do a test-run without actually deleting anything, run the script with the
Run in "live" mode
To call the script use the command
ephemetoot without any other arguments:
Depending on how many toots you have and how long you want to keep them, it may take a minute or two before you see any results.
Specify the config location (--config)
By default ephemetoot expects there to be a config file called
config.yaml in the directory from where you run the
ephemetoot command. If you want to call it from elsewhere (e.g. when using
cron), you need to specify where your config file is:
ephemetoot --config '~/directory/subdirectory/config.yaml'
Slow down deletes to match API limit (--pace)
--pace flag, delete actions are slowed so that the API limit is never reached, essentially borrowing the 'pace' method from the
Mastodon.py module. This is recommended for your first run, as unless you have tooted fewer than 30 times you are guaranteed to hit the API limit for deletions the first time you run
ephemetoot. If you do not toot very often on most days, it is probably more efficient to use the default behaviour for daily runs after the first time, but you can use
--pace every time if you prefer.
Increase the time between retry attempts when encountering errors (--retry-mins)
--retry-mins to increase the period between attempts to retry deletion after an error. The default value is one (1) minute, but you can make it anything you like. This is useful if your mastodon server is unreliable or frequently in "maintenance mode".
ephemetoot will make four additional attempts if it encounters an error, so the following command, for example, would wait 20 minutes between each retry, allowing the script to continue if there is an outage of 79 minutes or fewer:
ephemetoot --retry-mins 20
Include datestamp with every action (--datestamp)
If you want to know exactly when each delete action occured, you can use the
--datestamp flag to add a datestamp to the log output. This is useful when using
--pace so you can see the rate you have been slowed down to.
Include full error messages (--verbose)
Sometimes you might get an error and want to know more about what's triggering it. Use the
--verbose flag to print the full error to the console, instead of just the friendly version.
Hide skipped items (--hide-skipped)
If you skip a lot of items (e.g. you skip direct messages) it may clutter your log file to list these every time you run the script. You can suppress them from the output by using the
Hide everything (--quiet)
-q flag to suppress all logging except for the account name being checked and the number of toots deleted. Use the
-qqq flag will suppress all output. Exception messages will not be suppressed, other than
IndexError when any account has no toots to check.
Only archive deleted toots (--archive-deleted)
If you provide a value for
archive in your config file, the default is that all toots will be archived in that location, regardless of whether or not it is being deleted. i.e. it will create a local archive of your entire toot history. If you run ephemetoot with the
--test flag, this allows you to use create an archive without even deleting anything.
You can use the
--archive-deleted flag to only archive deleted toots instead.
Combining flag options
You can use several flags together:
ephemetoot --config 'directory/config.yaml' --test --hide-skipped
Use them in any order:
ephemetoot --pace --retry-mins 5 --datestamp --config 'directory/config.yaml'
Deleting old toots daily is the best approach to keeping your timeline clean and avoiding problems with the API rate limit.
Linux and FreeBSD/Unix
To run automatically every day on a n*x server you could try using crontab:
- enter a new line:
@daily /path/to/ephemetoot --config /path/to/ephemetoot/config.yaml
- exit with
Ctrl + x(nano)
On MacOS you can use the
--schedule flag to schedule a daily job with launchd. Note that this feature has not been widely tested so please log an issue if you notice anything go wrong.
Run from within your
or from anywhere else run:
ephemetoot --schedule directory
directory is where you installed
For example if
ephemetoot is saved to
ephemetoot --schedule /User/hugh/python/ephemetoot
ephemetoot will run at 9am every day (as long as your machine is logged in and connected to the internet). You can change the time it is scheduled to run, using the
--time flag with
ephemetoot --schedule [directory] --time hour minute
For example to run at 2.25pm every day:
ephemetoot --schedule --time 14 25