coursier cache action

A GitHub action to save / restore the coursier / sbt / mill / Ammonite caches of your build.

Usage

Add a coursier/cache-action@v7 step to your YAML workflow, like

    steps:
      - uses: actions/checkout@v5
      - uses: coursier/cache-action@v7

Cached directories

Coursier cache

Always cached.

Add files to take into account in its cache key via extraFiles.

~/.sbt and ~/.ivy2/cache

Cached when sbt files are found (any of *.sbt, project/**.scala, project/**.sbt, project/build.properties).

Add files to take into account in its cache key via extraSbtFiles.

~/.cache/mill

Cached when mill files are found (any of .mill-version, ./mill).

Add files to take into account in its cache key via extraMillFiles.

~/.ammonite

Cached when Ammonite scripts are found (any of *.sc, */*.sc).

Add files to take into account in its cache key via ammoniteScripts.

Parameters

root

Optional Root directory containing build definition sources (build.sbt, build.sc, etc.)

If the sbt or mill build definition files are in a sub-directory, pass the path to this sub-directory here.

path

Optional Override for the path of the coursier cache.

By default, the coursier cache is assumed to be in the default OS-dependent location. Set this input to override that. Note that this action will also set the COURSIER_CACHE environment variable if an override is specified, so that you don't have to set it yourself.

extraFiles

Optional Extra files to take into account in the cache key.

By default, sbt build definition files (*.sbt, project/**.{scala,sbt}, project/build.properties) and mill build definition files (*.sc, ./mill) are hashed to uniquely identify the cached data. Upon cache restoration, if an exact match is found, the cache is not saved again at the end of the job. In case of no exact match, it is assumed new files may have been fetched; the previous cache for the current OS, if any, is restored, but a new cache is persisted with a new key at the end of the job.

To take into account extra files in the cache key, pass via extraFiles either

• a single path as a string
• multiple paths in a JSON array, encoded in a string

Blobs are accepted (processed by @actions/glob).

extraSbtFiles

Optional Extra sbt files to take into account in the sbt cache key. Same format as extraFiles.

extraMillFiles

Optional Extra mill files to take into account in the mill cache key. Same format as extraFiles.

ammoniteScripts

Optional Extra Ammonite scripts to take into account in the Ammonite cache key. Same format as extraFiles.

extraKey

Optional

Extra value to be appended to the coursier cache key.

See extraFiles for more details.

extraHashedContent

Optional

Extra content to take into account in the cache key.

See extraFiles for more details.

The content of extraHashedContent is taken into account in the hash for the coursier cache key.

ignoreJob

Optional

Default: false

Set true if you don't want to use a job id as part of cache key.

ignoreMatrix

Optional

Default: false

Set true if you don't want to use a matrix jobs as part of cache key.

ignoreAmmonite

Optional

Default: false

Set true to skip saving and restoring the Ammonite cache, regardless of whether the repository contains .sc scripts.

disableFallback

Optional

Default: false

Set true to disable falling back to a less specific cache key when there is no exact cache hit. By default, if no exact cache match is found, the action will restore from a more general cache key (e.g. from a different job or build configuration). Disabling this prevents unintended cache sharing across jobs or matrix instances that have different dependency sets.

Cache invalidation

To manually invalidate the cache without changing your build files, set the COURSIER_CACHE_ACTION_CACHE_VERSION environment variable in your workflow. Changing its value produces a different cache key across all cache types, so the old cache is ignored and a fresh one is created.

The variable can be set at the workflow level so that it applies to every step:

env:
  COURSIER_CACHE_ACTION_CACHE_VERSION: 1   # increment this value to invalidate the cache

steps:
  - uses: actions/checkout@v5
  - uses: coursier/cache-action@v7

Or it can be scoped to only the coursier/cache-action step:

steps:
  - uses: actions/checkout@v5
  - uses: coursier/cache-action@v7
    env:
      COURSIER_CACHE_ACTION_CACHE_VERSION: 1   # increment this value to invalidate the cache

This is equivalent to passing the version as extraHashedContent, but more convenient because it can be set globally for the whole workflow (or even as a repository variable / secret) without touching the individual step configuration.

Outputs

• cache-hit-coursier - A boolean value to indicate a match was found for the coursier cache
• cache-hit-sbt-ivy2-cache - A boolean value to indicate a match was found for the sbt-ivy2-cache cache
• cache-hit-mill - A boolean value to indicate a match was found for the mill cache
• cache-hit-ammonite - A boolean value to indicate a match was found for the ammonite cache

See Skipping steps based on cache-hit for info on using this output

Skipping steps based on cache-hit

Using the cache-hit-... outputs above, subsequent steps can be skipped when a cache hit occurs on a given key.

Example:

steps:
  - uses: actions/checkout@v5

  - uses: coursier/cache-action@v7
    id: coursier-cache

  - name: Fetch Dependencies
    if: steps.coursier-cache.outputs.cache-hit-coursier != 'true'
    run: sbt +update

Note: The id defined in coursier/cache-action must match the id in the if statement (i.e. steps.[ID].outputs.cache-hit-coursier)