Claude Code now builds live dashboards for your whole team

A developer tool that once produced static code outputs can now feed live data directly into shared dashboards. On 18 July 2026, Anthropic updated Claude Code with a set of changes that shift it from a personal coding assistant into a small-team collaboration platform. The update matters most for developers at UK SMEs who are building internal tools on a tight budget and a tighter timeline.

What Anthropic changed in Claude Code

The headline addition is live data integration inside shared artefacts. Using Model Context Protocol connectors, developers can now pull real-time data feeds into the outputs Claude Code produces. A shared artefact that previously showed a static code snippet can now display a live chart, a refreshing table, or an updating metric.

Before this update, keeping a dashboard current meant a developer manually refreshing the data source or writing a separate sync script. That extra step added time and introduced the risk of stale figures reaching non-technical colleagues.

The release also adds public sharing links and collaborative editor roles for Team plan users. Non-technical staff can view or contribute to a shared artefact without needing their own developer environment. A terminal screen reader mode improves accessibility for assistive technology users. Improved background task handling means long-running AI commands no longer lock up the system mid-task. The full changelog is available in the Claude Code release notes.

Why this matters for UK software teams

For developers at UK SMEs, the previous workflow had a hidden cost. Every time a stakeholder needed fresh data in a shared output, a developer had to step back in and update it. That is fine once. It becomes a drain when it happens several times a week across multiple internal tools.

The live data connection removes that loop. A developer builds the artefact once, connects the MCP data source, and the output stays current without further input. That frees developer time for building new tools rather than maintaining old ones.

Collaborative editor roles on Team plans are equally significant. UK small technical teams often operate without the budget for enterprise-tier licences. Having shared editing available at the standard Team plan level means a two-person or three-person development function can work on the same artefact simultaneously. Non-technical colleagues in operations, finance, or sales can view live outputs in context rather than waiting for a PDF or screenshot.

gecco's Automations practice focuses on connecting live data feeds to shared outputs and removing the manual syncing steps that slow down small technical teams. That work maps directly to what this update makes easier inside Claude Code.

Three actions developers can take this week

1. Test MCP live data in one existing artefact. Pick an internal report your team already shares. Connect a live data source using an MCP connector and replace the static figures. This quick win takes under 30 minutes if the data source is already accessible via an API. You need an active Claude Code account to start.

2. Set up collaborative editor roles for your Team plan. If your team is on a Claude Code Team plan, review the new role options in account settings. Assign view access to at least one non-technical colleague who regularly requests data updates. Ask them to access the artefact directly rather than requesting a manual refresh from you. This removes one class of interruption from your week.

3. Audit which internal tools are still relying on manual data refreshes. List every dashboard, report, or shared output your team maintains that requires a developer to update the data by hand. That list is the starting point for migrating to live MCP connections. Prioritise by frequency: the tools that get updated most often give the fastest return on migration effort.

Considerations and limitations

MCP connectors require the data source to expose an accessible endpoint. If your internal systems sit behind a firewall or use proprietary formats, connecting them to Claude Code will need additional configuration work. That is not a blocker, but it is planning time that should not be underestimated.

Collaborative editor roles are available on Team plans, but the specific permissions and role tiers are worth checking against your current licence before rolling out to non-technical colleagues. Anthropic may adjust these as the feature matures.

Public sharing links carry a straightforward risk: an artefact shared publicly is visible to anyone with the link. UK software teams should confirm what data a shared artefact exposes before enabling public links, particularly where the live data source contains commercially sensitive figures or personally identifiable information.

The background task improvements reduce lockups but do not eliminate them. Complex or poorly scoped commands can still cause delays. Teams adopting this for production workflows should test thoroughly before relying on it for time-sensitive outputs.

Where to take this next

If your development team is weighing up which of these new features to prioritise, and whether the broader shift toward live, collaborative AI tooling fits your current setup, that is exactly the kind of question the AI Readiness survey is built to surface. Taking the survey gives you access to 65 or more free resources and a custom AI Readiness report. That report leads directly to a free 45-minute AI Readiness call to walk through the results with you.

gecco's Training and consultancy programme helps UK software teams and developers build the internal capability to own tools like Claude Code rather than depend on outside support. Public sharing links and collaborative editor roles on Claude Code's Team plans open real possibilities for non-technical staff, but only if SMEs establish clear handoff protocols and upskill their teams to own these tools rather than outsource them.

Website · LinkedIn · Case Studies · Newsletter