Use Prismic

Prismic MCP Server

Last updated

The Prismic MCP server lets AI assistants work directly with your Prismic content. You can search documents, read and create content, upload assets, and manage releases from a conversation with Claude or ChatGPT.

Activate Prismic MCP

Before connecting Claude or ChatGPT, activate Prismic MCP on each repository you want your assistant to use.

Connect to the MCP server

The MCP server URL is:

https://mcp.prismic.io/mcp

After activation, add this URL to your AI assistant.

Authentication

When you first use the Prismic MCP server, your AI assistant will prompt you to authenticate. Click the authentication button to be redirected to Prismic’s login page. Sign in with your Prismic account and the connection will be established.

The MCP server uses your Prismic identity, so you only have access to repositories your account can access.

Available tools

The MCP server exposes the following tools. Your AI assistant calls these automatically based on your prompts.

Discovery

ToolDescription
list_repositoriesList all Prismic repositories you have access to. Called first to discover available repositories.
list_custom_typesList all custom types (content models) in a repository with their id, label, and format.
list_shared_slicesList all shared slices in a repository with their variation names.
list_localesList all locales configured in a repository.
list_releasesList all releases in a repository, newest first.
list_document_versionsList every version of a document, including drafts, published, archived, and release versions.

Read

ToolDescription
search_documentsSearch documents by keywords with optional filters for custom type, locale, status, and tags. Returns metadata, not full content.
search_assetsSearch the asset library by keyword. A single query matches against filename, alt text, notes, and credits. Returns previews.
get_documentFetch a document’s metadata and full content. Supports reading from a specific version or release.
get_custom_typeFetch a custom type’s full field model and an empty content template for creating documents.
get_shared_sliceFetch a shared slice’s full model with all variations and ready-to-use empty variation bodies.

Write

ToolDescription
create_releaseCreate a new empty release. Returns its id for use with create_document and update_document.
create_documentCreate a new document inside a release. Missing fields are filled with model defaults.
update_documentUpdate a document into a release with full content. The new version is branched from a base version.
upload_assetFetch an asset from a URL and upload it to the repository’s asset library. Returns the Prismic CDN url for use in documents.

Example prompts

You can use the following prompts once connected to Prismic.

Explore your content

Prompt
What repositories do I have access to in Prismic?
Prompt
Show me the content types and locales configured in my-website.

Search and read

Prompt
Find all blog posts about product updates in my-website.
Prompt
Show me the content of the homepage.

Create and edit content

Prompt
Create a new blog post about our latest feature release.
Prompt
On all blog posts about ABM, update the CTA card to promote our new feature.
Make each one relevant to the article.
Prompt
Rewrite the hero description on the pricing page to be more concise.

The AI assistant reads the document, proposes changes, and saves everything to a release in Prismic for you to review.

Troubleshooting

The connector isn’t working

If the Prismic connector stops working — for example, tools fail with authentication errors, or it stops responding after working previously — and removing and re-adding it doesn’t fix the problem, your Prismic authentication session may be stuck. Your AI assistant reuses that session each time you reconnect, so the bad state persists.

To reset it:

Was this page helpful?