Installing the Adology Content Intelligence Plugin

This guide walks you through installing the Adology Content Intelligence plugin in your AI agent of choice. Three install paths cover the supported clients:

Claude.ai connector — zero-install, works in the Claude.ai web app. Recommended for most customers.
Claude Code plugin — for developers and power users running Claude Code in a terminal.
Claude Cowork plugin — for Claude Desktop users on macOS or Windows running the Cowork agentic workspace.

All three paths use the same hosted Adology MCP server (https://mcp.adologyai.com/mcp) and the same Adology account.

Capability comparison

Capability	Claude.ai connector	Claude Code plugin	Claude Cowork plugin
43 MCP tools (`whoami`, portfolios, projects, search, analyze, …)	✅	✅	✅
11 domain skills (`brand-builder`, `audience-deep-dive`, …)	—	✅	✅
4 slash commands	—	✅	✅
OAuth via Stytch	✅	✅	✅
Runs on	Web (any browser)	macOS / Linux / Windows terminal	Claude Desktop on macOS / Windows

Claude.ai gives you the MCP tools but not the bundled skills and slash commands. Claude Code and Cowork install the full plugin bundle.

Before you begin

You need an Adology account to use the plugin.

Go to https://adologyai.com and sign up (or log in if you already have an account).
Keep your login credentials handy — you’ll sign in to Adology during the OAuth flow when your AI agent first connects.

Option A: Claude.ai connector (recommended for most users)

The Adology MCP server is a hosted HTTPS service with OAuth 2.1 authentication, so Claude.ai can connect to it as a custom connector. No local installation required.

Step 1 — Open Claude.ai connector settings

Sign in to https://claude.ai.
Click your profile picture (top-right) → Settings.
Open the Connectors tab.

Step 2 — Add the Adology connector

Click Add custom connector.
Fill in:
- Name: Adology
- URL: https://mcp.adologyai.com/mcp
Click Connect.

Claude.ai will fetch the OAuth metadata from the server and redirect you to the Adology login flow (powered by Stytch).

Step 3 — Authenticate

Sign in with your Adology credentials.
Approve the requested scopes (workspace read, scrape/analysis triggers, and writes for items you explicitly save).
You’ll be redirected back to Claude.ai. The connector should now show as Connected.

Step 4 — Verify

Open a new chat in Claude.ai and ask:

Use the Adology whoami tool and tell me which workspace I’m connected to.

Claude should call the tool and reply with your workspace details. You’re done.

Option B: Claude Code plugin

For developers using Claude Code in a terminal. The recommended path installs directly from GitHub — no local clone required.

Step 1 — Confirm prerequisites

Claude Code is installed and up to date. Check with:
```
claude --version
```
Install or upgrade from https://docs.claude.com/en/docs/claude-code/quickstart.
You have an Adology account (see Before you begin). No API token to manage — the plugin uses OAuth 2.1.

Step 2 — Install from GitHub

Start Claude Code in any directory:

claude

Inside the Claude Code session, add the Adology marketplace using GitHub’s owner/repo shorthand and install the plugin:

/plugin marketplace add adologyai/content-intelligence-plugin
/plugin install content-intelligence@adology-marketplace

Claude Code fetches the marketplace catalogue and plugin contents directly from GitHub. No clone, no local path management.

Private network or air-gapped environment? See Offline install below.

Step 3 — Authenticate (first tool call)

The first time Claude Code calls an Adology tool, it will start the OAuth 2.1 flow against the hosted MCP server. Run /mcp to inspect connector status or trigger auth manually — Claude Code will open a browser tab, you sign in with your Adology credentials (powered by Stytch), and the token is stored locally for future sessions.

Step 4 — Verify

In the same Claude Code session, list installed plugins:

/plugin list

You should see Adology — Content Intelligence. Then ask the agent:

Use the Adology whoami tool.

If OAuth hasn’t completed yet, Claude Code prompts you through the sign-in flow. After that, the tool returns your workspace info. You’re done.

Option C: Claude Cowork plugin

For Claude Cowork users running Claude Desktop on macOS or Windows. Cowork uses Claude Desktop’s built-in plugin manager rather than a CLI command.

Step 1 — Confirm prerequisites

Claude Desktop for macOS or Windows. Install from https://claude.ai/download.
Cowork access on your Claude account (Pro or Max). Cowork is desktop-only; it does not run on web or mobile.
You have an Adology account (see Before you begin). No API token to manage — the plugin uses OAuth 2.1.

Step 2 — Install from the plugin directory

Open Claude Desktop and switch to the Cowork tab.
Click Customize in the left sidebar.
Click Browse plugins.
Search for Adology or content-intelligence and click Install on the Adology Content Intelligence plugin.

Plugin not yet in the directory? While the listing is pending approval, install manually instead:

Download content-intelligence-<version>.zip from https://github.com/adologyai/content-intelligence-plugin/releases/latest.

In Claude Desktop → Cowork → Customize, choose the option to upload a custom plugin file and select the downloaded zip.

Step 3 — Authenticate (first tool call)

The first time Cowork calls an Adology tool, it kicks off the OAuth 2.1 flow against the hosted MCP server. A browser tab opens for you to sign in with your Adology credentials (powered by Stytch); the token is stored locally for future sessions.

Step 4 — Verify

In a Cowork conversation, ask:

Use the Adology whoami tool and tell me which workspace I’m connected to.

You should see your workspace info returned.

Offline install (Claude Code)

If your environment can’t reach GitHub from inside Claude Code (e.g., corporate proxy, air-gapped network), clone the repository on a machine that has GitHub access and point the marketplace at the local copy instead.

git clone https://github.com/adologyai/content-intelligence-plugin.git ~/.adology/content-intelligence-plugin

Then in a Claude Code session:

/plugin marketplace add ~/.adology/content-intelligence-plugin
/plugin install content-intelligence@adology-marketplace

Use the absolute path if ~ does not expand for you.

Updating the plugin

Claude.ai connector

Nothing to do — the server is hosted by Adology and updates automatically.

Claude Code plugin (GitHub install)

In a Claude Code session:

/plugin marketplace update adology-marketplace
/plugin install content-intelligence@adology-marketplace

/plugin marketplace update refreshes the catalogue from GitHub; the second command installs the latest version.

Claude Code plugin (offline install)

Pull the latest changes locally, then refresh:

cd ~/.adology/content-intelligence-plugin
git pull

/plugin marketplace update adology-marketplace
/plugin install content-intelligence@adology-marketplace

Claude Cowork plugin

If you installed via Browse plugins, Claude Desktop auto-updates the plugin to the latest version published in the directory — no action needed.

If you installed via upload a custom plugin file, download the new content-intelligence-<version>.zip from https://github.com/adologyai/content-intelligence-plugin/releases/latest and re-upload it via Claude Desktop → Cowork → Customize.

Troubleshooting

`401 Unauthorized` or “authentication failed”

All three clients use OAuth 2.1, so the recovery is the same: re-run the OAuth flow to get a fresh access token.

Claude.ai connector: open Settings → Connectors, disconnect Adology, and reconnect.
Claude Code: run /mcp in a session and re-authenticate the adology server. If the issue persists, fully quit Claude Code, relaunch, and try again — stale tokens can survive an in-session reconnect.
Claude Cowork: in Claude Desktop → Cowork → Customize, disable and re-enable the Adology plugin (or uninstall and reinstall). The next tool call triggers a fresh OAuth flow.

Plugin doesn’t appear in `/plugin list`

Confirm the marketplace was added:

/plugin marketplace list

If adology-marketplace isn’t listed, repeat Step 2 (or the Offline install steps, if you went that route).

MCP server unreachable

Check connectivity to the hosted server:

curl -sS https://mcp.adologyai.com/.well-known/oauth-protected-resource

You should get back a JSON document. If this fails, the issue is network-level (firewall, VPN, DNS) rather than the plugin.

Empty results from analysis tools

New data pulls need a few minutes to land. Ask the agent to call check_pull for the run — fetches typically take 5–15 minutes.

Getting help

Email: hello@adologyai.com
Bug reports: https://github.com/adologyai/content-intelligence-plugin/issues
Full support guide: SUPPORT.md

We aim to respond within one business day.