How to Use AI Agents to Manage Your Game Asset Catalog

Your AI assistant can use your tools

AI assistants are no longer limited to answering questions from their training data. With the Model Context Protocol (MCP), assistants like Claude can connect to external tools and perform real operations — reading data, creating records, and executing searches with the same permissions as a human user.

PolyDrobe ships with a built-in MCP server that exposes the core asset management operations. This article walks through connecting an AI assistant to your PolyDrobe project and shows practical workflows you can run today.

Setting up the connection

What you need

• A PolyDrobe account with at least one project
• An AI assistant that supports MCP (Claude Desktop, Claude Code, or any MCP-compatible client)
• A Bearer token for authentication (obtained via PolyDrobe's OAuth endpoint at /o/token/)

Connecting Claude

In your Claude configuration, add PolyDrobe as an MCP server:

{
  "mcpServers": {
    "polydrobe": {
      "url": "https://polydrobe.app/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_TOKEN_HERE"
      }
    }
  }
}

Once connected, Claude automatically discovers the available tools and can start using them in conversation. You do not need to tell Claude what tools exist — it reads the tool descriptions from the MCP server and decides which ones to use based on your request.

Practical workflows

Here are real workflows you can run once connected. Each example shows what you say to the assistant and what happens behind the scenes.

Browse your catalog

You say:

"List all my projects and show me what is in the Vanguard Skins project."

What happens: The assistant calls list_projects to find your projects, identifies "Vanguard Skins" by name, then calls get_project to retrieve the full project details including categories, rarities, statuses, and summary statistics. You get a structured overview without opening the browser.

Search for specific assets

You say:

"Find all helmets with Epic rarity in the Vanguard Skins project."

What happens: The assistant calls list_assets with category and rarity filters, or uses the search tool with a combined query. It returns matching assets with their variant counts, statuses, and key metadata.

This is especially useful during planning meetings when someone asks "how many Epic helmets do we have?" and you need the answer in seconds, not after navigating to the project, setting filters, and counting rows.

Create assets and variants

You say:

"Create a new asset called Phantom Jacket in the Outerwear category with Rare rarity. Add two variants: Phantom Jacket — Obsidian at 800 coins and Phantom Jacket — Chrome at 1200 coins with Epic rarity."

What happens: The assistant calls create_asset with the name, category, and rarity, then calls create_variant twice with the correct parent asset, names, prices, and rarity values. Three API calls, one sentence from you.

This workflow shines during brainstorming sessions. Instead of switching to the browser, clicking through forms, and losing your train of thought, you describe the asset in natural language and it appears in the catalog.

Check release readiness

You say:

"Give me a release readiness report for Release 2.1. How many variants are Approved, how many are still In Progress, and which ones are blocked in Review?"

What happens: The assistant calls list_assets filtered by release version, iterates through variants, groups them by status, and presents a summary:

• 47 variants Approved
• 12 variants In Progress
• 3 variants in Review (lists names and parent assets)

This replaces manual filtering and counting in the web interface. During a release planning standup, you can get an up-to-the-minute status check by asking a question.

Compare variants

You say:

"Show me all Legendary-rarity variants across the project with their prices. Are any priced below 2000 coins?"

What happens: The assistant queries the catalog, filters by rarity, extracts pricing from each variant, and flags any that fall below the threshold. This is a pricing consistency check that would normally require exporting data and running a spreadsheet formula.

Generate documentation

You say:

"Write a summary of all Season 3 Battle Pass items including asset names, variant counts, rarity distribution, and total item count."

What happens: The assistant searches for items tagged "Season 3" or "Battle Pass", aggregates the data, and produces a formatted summary you can paste into a design document, Slack message, or email to a publisher.

Bulk edits across many variants

You say:

"Create ten color variants of the Phantom Jacket — Sapphire, Ruby, Emerald, Obsidian, Ivory, Bronze, Copper, Jade, Violet, and Amber — all priced at 800 coins with Rare rarity."

What happens: The assistant calls bulk_create_variants once with the full list, so ten variants land in the catalog in a single round trip. The same toolset exposes bulk_update_variants and bulk_delete_variants for mass status changes, repricing, or cleanup.

Manage tags, categories, and releases

You say:

"Add a new category called Cyber Ops under Outerwear, create a Legendary rarity with the color #ffb347, and tag every Phantom Jacket variant with the Season 4 tag."

What happens: The assistant calls create_category, create_rarity, then add_variant_tags against the matching variants. Classification tools cover categories, tags, rarities, releases, currencies, and asset statuses — the same vocabulary you manage in project settings, now available to your AI assistant.

Export and import catalog data

You say:

"Export the Vanguard Skins project as JSON so I can review it offline, then re-import the edited file when I am done."

What happens: The assistant calls export_project to retrieve a JSON snapshot of the project and import_project_data to apply an edited payload back. This is useful for data migrations, offline editing, or moving a curated subset of assets between projects.

What the assistant cannot do

MCP respects the same permission model as the web interface. A Viewer's token can read and search but cannot create or modify data. An Editor's token allows creation and updates on assets, variants, classifications, and tags. An Owner's token additionally unlocks team member management, project deletion, and destructive operations.

The MCP server does not currently support:

• Uploading full-resolution 3D model files (variant thumbnails can be set via set_variant_thumbnail, but textures and mesh files are managed through the web interface)
• Authentication and account management (sign-up, password reset, billing actions)
• Stripe or workspace billing changes

Everything else that exists in the PolyDrobe web interface — full CRUD on projects, assets, and variants, classification management, tag assignment, member roles, activity logs, and bulk operations — is exposed through MCP tools.

Building automation on top of MCP

Beyond interactive use with Claude or ChatGPT, the MCP server enables programmatic automation:

Pre-release validation. A CI pipeline step that connects to the MCP server, queries all variants targeted for the next release, and fails the build if any are still in "In Progress" status.

Slack integration. A bot that responds to "What shipped in Release 2.0?" by querying the catalog and posting a formatted summary to the channel.

Content planning. A script that pulls the current catalog state, compares it against a target (e.g., "every category should have at least 3 Legendary variants"), and generates a gap report.

Cross-tool sync. An agent that reads asset data from PolyDrobe and creates corresponding Jira tickets for production tasks, keeping the asset catalog and task tracker in sync.

Getting started

1. Log in to PolyDrobe and make sure you have at least one project with some assets.
2. Upgrade to a plan that includes MCP access (Professional or Enterprise — MCP is not enabled on the free tier).
3. Generate a Bearer token via the OAuth endpoint.
4. Add the MCP server to your AI assistant's configuration.
5. Start with a simple query: "List my projects" — and go from there.

If your team is evaluating whether AI-assisted asset management fits your workflow, the pricing page lists which plans include MCP access alongside the rest of the feature matrix.

Key takeaways

• PolyDrobe's MCP server lets AI assistants browse, create, update, and delete game assets, variants, and classifications using natural language.
• Practical workflows include catalog queries, batch asset creation, release readiness checks, pricing audits, bulk variant edits, and data export/import.
• Authentication uses OAuth 2.1 with the same role-based access control as the web interface — Viewer, Editor, and Owner tokens behave as they do in the UI.
• The MCP server enables automation beyond interactive use: CI validation, Slack bots, content planning, and cross-tool sync.
• MCP access is available on Professional and Enterprise plans — connect any MCP-compatible AI assistant to start using it.