Acquia Source logo

Acquia Source

Community
grasmash
acquia-source

Connect to an Acquia Source ("Source powered by Drupal CMS") site — its JSON:API and MCP endpoints, OAuth2 auth, the @drupal-canvas CLI, and the Canvas Page API deploy workflow. Use when wiring an AI agent or app to a Source CMS site, building/uploading Drupal Canvas Code Components, deploying landing pages via the Page API, or registering the Source MCP server with Claude Code/Cursor.

Overview

Publishergrasmash
Repositorydrupal-claude-skills
Skill nameacquia-source
Stars
77
Forks
4
Bundled files
Instructions only
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by grasmash on GitHub. Read the source before you install it.

Installation

Install the Acquia Source AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/grasmash/drupal-claude-skills.git /tmp/drupal-claude-skills
mkdir -p .claude/skills
cp -r /tmp/drupal-claude-skills/skills/acquia-source .claude/skills/acquia-source
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Acquia Source in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Acquia Source on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Acquia Source is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Acquia Source

Acquia Source ("Source powered by Drupal CMS") is a managed SaaS CMS wrapping Drupal CMS 11+ and the Drupal Canvas page builder (drupal.org/project/canvas, formerly Experience Builder). You don't get server/SSH access or arbitrary contrib — you interact through its JSON:API, its MCP server, the Canvas Page API, and the @drupal-canvas CLI. Admin/login at https://source.acquia.com.

Acquia Source is a SaaS product, not Acquia Cloud Platform. The acli CLI manages Cloud Platform, not Source sites — don't reach for it here. Source is driven through the web admin + the APIs below.

Site identity & credentials

  • CMS URL looks like https://<uuid>.cms.acquia.site (auto-generated at site creation; a custom domain can be mapped later). This base URL is the JSON:API host and the OAuth host.
  • API clients (OAuth2 client_id/secret) are created in the admin UI under API → API clients → Add API client.
  • The @drupal-canvas CLI reads everything from a local .env (never committed):
    CANVAS_SITE_URL=https://<uuid>.cms.acquia.site
    CANVAS_CLIENT_ID=<client_id>
    CANVAS_CLIENT_SECRET=<client_secret>
    CANVAS_JSONAPI_PREFIX=api        # Source serves JSON:API under /api, not the Drupal default /jsonapi
    # CANVAS_ACCESS_TOKEN=<token>    # optional: a pre-issued Bearer token instead of client credentials

Authentication (OAuth2 via simple_oauth)

  • Token endpoint: POST https://<site>/oauth/token, Content-Type: application/x-www-form-urlencoded.
  • Grants: Authorization Code (user-present), Client Credentials (server/agent — use this for automation), Refresh Token.
  • Request header on API calls: Authorization: Bearer <access_token>.
  • Scopes: content scopes like content:read / content:write (examples — the full list isn't published); the Canvas CLI needs canvas:asset_library and canvas:js_component. Enable Client Credentials + add those scopes when creating the API client.

Client-credentials token, by hand:

bash
curl -s -X POST "$CANVAS_SITE_URL/oauth/token" \
  -d grant_type=client_credentials \
  -d client_id="$CANVAS_CLIENT_ID" \
  -d client_secret="$CANVAS_CLIENT_SECRET" \
  -d scope="canvas:asset_library canvas:js_component" | jq -r .access_token

JSON:API

  • Base path is /api on Source (Drupal's default /jsonapi is overridden — set CANVAS_JSONAPI_PREFIX=api). So a collection is https://<site>/api/<entity_type>/<bundle> (e.g. /api/node/article).
  • Resource types are standard Drupal {entity_type}--{bundle} (node--article, media--image, taxonomy_term--tags, …). No Source-proprietary prefix confirmed.
  • The admin UI has a JSON:API Query Builder (API → JSON:API Query Builder) that generates live queries + code, and per-site OpenAPI documentation (API → OpenAPI documentation). Use those to discover exact bundles/fields for a given site.
bash
curl -s -H "Authorization: Bearer $TOKEN" \
  "$CANVAS_SITE_URL/api/node/article?page[limit]=5" | jq '.data[].attributes.title'

MCP server (connect an AI agent)

Acquia Source ships an MCP server (HTTP transport, OAuth2 with dynamic client registration). It's experimental and per-site — enable it first.

  1. Enable: admin UI → Configurations → Experimental features → MCP Server toggle.
  2. Get the URL: copy the site's MCP URL from MCP Client Configuration in the Source admin (the endpoint URL is provisioned per site; there's no fixed public path to hardcode).
  3. Register in Claude Code:
    bash
    claude mcp add acquia-source-mcp <ACQUIA_SOURCE_SITE_MCP_URL>
    Then run /mcp, select acquia-source-mcp, and click Authenticate to complete the OAuth flow (dynamic client registration — no manual client_id needed).

What it exposes (read-only resources + write tools, via drupal:// and canvas:// URIs):

  • Content Management — content types, media types, vocabularies, text formats; create/update nodes, media, taxonomy terms.
  • Content Modeling — field storages, field types; create content types and add fields.
  • Drupal Canvas — components and pages; manage layouts and component properties.

Prefer the MCP server over hand-rolled JSON:API calls when an agent needs to model content or edit Canvas layouts conversationally; use JSON:API/the Page API for deterministic, scripted deploys.

Canvas Code Components — CLI workflow

Scaffold and manage React/JSX Code Components with the official tooling (install the official Canvas skills for the component-authoring rules — don't reinvent them):

  • npx @drupal-canvas/create@latest — scaffold a codebase (default template: acquia/nebula, which ships AGENTS.md + SETUP.md with the Page API reference and pitfalls).
  • npx canvas download / npx canvas upload -c <comp1,comp2> -y — pull/push components between repo and site. canvas.config.json (committed) defines structure; .env (uncommitted) holds the secrets above.

Canvas Page API — deploying a page

Pages are built by uploading components, then PATCHing the page with a full component tree. The hard-won rules (these cost hours otherwise):

  • Each PATCH replaces the ENTIRE page. Never build incrementally with multiple PATCHes — assemble the complete component tree and send it in one atomic request.
  • Version hashes change on every upload. Re-discover all component version hashes after each canvas upload; never hardcode them. Only changed components get new hashes.
  • FormattedText strips inline style attributes on the live site (works in Storybook, fails live). For text alignment add a layout prop (left_aligned/center_aligned/right_aligned) with CVA variants — never inline style="text-align:…".
  • The Image component needs Drupal media-pipeline URLs (with alternateWidths). External URLs / data URIs are accepted without error but render as invisible 0×0 images live. For logos use inline SVG/text; for content images upload as Drupal media first.
  • No raw hex colors as props — use CVA variants mapped to Tailwind theme tokens in global.css.
  • Padding stacks: section adds px-4, grid_container adds px-6 → ~80px on a 375px screen. Use responsive prefixes (px-0 md:px-6) on inner containers.
  • Empty slot containers need min-w-*/min-h-* so they stay interactive in the Canvas editor.
  • Deploy script pattern: read .env → client-credentials token → discover all version hashes in one pass → build the tree (parents before children, fresh UUID per instance, inputs as a json.dumps() string) → single PATCH. Keep a VERSIONS dict at the top.

Docs

Frequently asked questions

What does the Acquia Source AI skill do?

Connect to an Acquia Source ("Source powered by Drupal CMS") site — its JSON:API and MCP endpoints, OAuth2 auth, the @drupal-canvas CLI, and the Canvas Page API deploy workflow. Use when wiring an AI agent or app to a Source CMS site, building/uploading Drupal Canvas Code Components, deploying landing pages via the Page API, or registering the Source MCP server with Claude Code/Cursor.

Why use Acquia Source on TypingMind?

Because you install it once and use it with any model. Acquia Source is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Acquia Source in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/grasmash/drupal-claude-skills/tree/main/skills/acquia-source. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Acquia Source?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Acquia Source?

As many as you like. As long as a model supports skills, you can use Acquia Source with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Acquia Source AI skill free?

It is published on GitHub by grasmash. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇