DID-ACTO / FIELD GUIDE

Connect Claude or Codex

Use the public MCP endpoint with your own DiD-Acto identity.

The connection details

The same server endpoint works for both configured clients:

https://prolug.asuscomm.com/mcp

Client	OAuth client ID	Client secret
Codex	`didacto-codex`	Leave unset
Claude remote connector	`didacto-claude`	Leave unset

These client IDs identify the application; they are not passwords. Sign in through the DiD-Acto page. Never paste your password or a bearer token into the research conversation.

Codex: complete configuration

Add this entry to your Codex config.toml. If prolug already exists, update that entry instead of creating a duplicate. Keep other MCP entries unchanged.

[mcp_servers.prolug]
url = "https://prolug.asuscomm.com/mcp"

[mcp_servers.prolug.oauth]
client_id = "didacto-codex"
callback_url = "http://127.0.0.1/callback"

Run codex mcp login prolug, complete the DiD-Acto sign-in, then restart Codex so future tasks can load the tools. The per-server callback avoids conflicts with a global localhost setting. Do not replace it with localhost.

This service uses pre-registered OAuth clients. Dynamic registration is not supported and is not advertised. The portal provides a Copy Codex configuration button. No password or bearer token needs to be pasted into a chat. See the official Codex MCP documentation.

Claude remote connector

Open Claude’s connector settings, usually under Customize → Connectors.
Add a custom connector with the MCP URL above.
In advanced settings, enter didacto-claude as the OAuth client ID. Leave the secret blank.
Select Connect and complete the DiD-Acto sign-in.
Enable the connector for the conversation in which you want to use it.

On an organizational account, an owner may need to add the connector first. See Anthropic’s custom connector instructions for the current menu labels and account requirements.

This configuration targets Claude’s remote connector, whose requests come through Anthropic’s infrastructure. Do not assume the same client registration applies to Claude Code or a local stdio setup.

Check the connection

Try asking your assistant

Check that DiD-Acto is connected. List its available workflow tools and resources. Do not upload data or start an analysis yet.

The workflow has six tools: register_dataset, inspect_dataset, plan_analysis, run_analysis, get_result and render_report. The connection also exposes three static reference resources. If tools are missing, check the connector is enabled and you signed in successfully.

Complete a first sample workflow

After connecting, ask your assistant to inspect a shared sample. This checks that the tools are available in your conversation and gives you a familiar starting point before uploading your study data.