dcc-asset-objaverse

DCC-MCP Objaverse Assets

Agent workflow

AI agents should use installed package skills through the shared gateway. IDE users may continue to use the MCP endpoint.

Install or update the CLI

dcc-mcp-cli is the preferred control path for every shell-capable agent. If it is missing, ask the user before installing the latest official release:

On Linux/macOS, download and inspect the installer:

curl -fL https://raw.githubusercontent.com/dcc-mcp/dcc-mcp-core/main/scripts/install-cli.sh -o install-cli.sh
cat install-cli.sh

After completing the inspection and obtaining the user's permission to install:

sh ./install-cli.sh

On Windows PowerShell, download and inspect the installer:

Invoke-WebRequest https://raw.githubusercontent.com/dcc-mcp/dcc-mcp-core/main/scripts/install-cli.ps1 -OutFile .\install-cli.ps1
Get-Content -Raw .\install-cli.ps1

After completing the inspection and obtaining the user's permission to install, run the reviewed local installer under the current execution policy:

& .\install-cli.ps1

Keep an official build current through the release manifest:

dcc-mcp-cli update check

Only after the user explicitly confirms the update:

dcc-mcp-cli update apply --yes

update apply downloads and stages the latest CLI for the next launch. It does not update a running dcc-mcp-server; update that server in its own environment.

Use --require-gateway for task calls so Gateway audit/stats records them. This fails closed instead of falling back to direct MCP. Choose one stable <task-id> and pass it to every tool call with --agent-session-id. This sets _meta.agent_context.session_id; reuse the same ID in stats --session-id after the task.

dcc-mcp-cli dcc-types
dcc-mcp-cli list
dcc-mcp-cli search --query "<task>" --dcc-type <host>
dcc-mcp-cli describe <tool-slug>
dcc-mcp-cli call <tool-slug> --require-gateway --agent-session-id "<task-id>" --json '{"key":"value"}'

If the package skill is not active, call dcc-mcp-cli load-skill <skill-name> --dcc-type <host>. After the task, query dcc-mcp-cli stats --range "<task-window>" --session-id "<task-id>" and pass only bounded evidence to the review_skill_improvement prompt from dcc-mcp-skills-creator. Choose a supported --range covering the full task start-to-review window; see dcc-mcp-cli stats --help.

Objaverse discovery and GLB download tools for DCC-MCP.

This skill is a thin wrapper around the official objaverse Python package. It does not import files into a DCC scene; host import belongs in host skills.

Install

dcc-mcp-cli marketplace add dcc-mcp/dcc-asset-objaverse
dcc-mcp-cli marketplace install dcc-asset-objaverse

Requirements

python -m pip install objaverse

Tools

• sample_objaverse_uids
• get_objaverse_annotations
• download_objaverse_objects

License Notes

This wrapper is MIT licensed. Objaverse as a dataset is licensed separately; individual objects carry their own Creative Commons licenses. Always keep the returned license and attribution metadata with downloaded assets.