Search, Pull, and Vet an Image
A 5-step recipe with Docker. We can help you get access.
Recipe · 5 steps
Search, Pull, and Vet an Image
with Docker
Search for an image, pull the best match, and vet its metadata and layers.
- Validate credentials against the registry before pulling. Returns 200 with an identity token or 204 when the credentials are good.
- Search the registry for the term, filtered so that only well-starred or official repositories come back. This is the step that turns a guessed image name into a vetted candidate.
- Pull the chosen image and tag into the local image store. Note that omitting a tag pulls every tag for the repository, so the tag is always supplied here.
- Inspect what actually arrived — the image id, repo digests, architecture, and the config the image will run with by default.
- Walk the image's parent layers to see how it was assembled. This is the supply-chain check: each entry carries the command that created the layer and its size.
What you'll need to have ready
- term: Term to search the registry for (e.g. "postgres").
- image: Name of the image to pull once the search has been reviewed.
- tag: Tag of the image to pull (e.g. "16-alpine").
- searchLimit: Maximum number of search results to return.
- searchFilters: JSON-encoded search filters, e.g. {"is-official":["true"],"stars":["100"]}.
- username: Registry username, when vetting against a private registry.
- password: Registry password or token.
- serveraddress: Registry address (e.g. "https://index.docker.io/v1/").
- registryAuth: Base64url-encoded auth configuration used for the pull.
For your developer
Arazzo workflow docker:search-and-pull-image, 5 operations: SystemAuth → ImageSearch → ImageCreate → ImageInspect → ImageHistory. Specs on APIs.io.