Field noteNov 21, 20265 min read1 source

Claude Code 2.1.330 adds Enterprise Admin Console sign-in and built-in charting mod

In one minute

Claude Code v2.1.330 adds Enterprise Admin Console sign-in, a built-in charting mod, and new controls and metrics for self-hosted runners. The release gives teams more ways to apply organization settings at login and inspect runner timing, while extending recurring tasks and /loop sessions to as long as 30 days.

What changed: Claude Code v2.1.330 adds sign-in for organizations managed in Anthropic’s Enterprise Admin Console and built-in charting through mods.

What happened

Anthropic’s Claude Code release notes identify v2.1.330 as a release published on November 20. The most visible addition for centrally managed teams is sign-in to organizations managed in Anthropic’s Enterprise Admin Console. After claude auth login, Claude Code applies the organization’s settings and uses its cloud provider or gateway. The notes do not describe the full policy surface or enumerate which settings are applied, but the change makes the organization-managed login path part of the CLI’s documented behavior.

The release also adds easel, a built-in mod that draws charts when asked and when Claude Code presents MCP metrics results. It is available on Pro and Max, and on Team and Enterprise where available, on macOS and Linux. Users can turn automatic charts off in /config. The release notes do not specify which chart formats or data sources easel supports beyond the named MCP metrics results.

For self-hosted runners, operators can use --start-dir or SELF_HOSTED_RUNNER_START_DIR to have Claude begin work in a folder of a repository. The notes say settings and hooks still load as they do without this option. The release adds two metrics to the runner’s /metrics endpoint: claude_code_self_hosted_runner_order_to_registered_seconds, measuring seconds from work order issuance to accepted registration on orchestrator-spawned runners, and claude_code_self_hosted_runner_order_to_first_session_seconds, measuring seconds from work order to the first session being handed over on runners started for a session.

The notes also change recurring scheduled tasks and /loop to run for up to 30 days while their session remains active, and to say when they end. Tasks created before this release still end after seven days. Other additions include a fuller sign-in link option through CLAUDE_CODE_FULL_SIGN_IN_LINK=1, more informative plugin installation JSON fields, and many fixes across terminal, MCP, Code Mode, Windows, VS Code, and self-hosted runner behavior.

What the archive adds

New Runtime’s earlier coverage of Claude Code v2.1.310 described Monitor support across Bedrock, Vertex, and Foundry, gateway retry, and Enterprise role controls (Claude Code v2.1.310, October 24, 2026). Version 2.1.330 is a new release on the same product line, with a distinct enterprise sign-in path and runner controls rather than a restatement of those earlier changes.

Trend signal

This release modestly supports agent-ready software exposes capabilities: organization-managed login and runner configuration are concrete control surfaces around how Claude Code connects and starts work. It does not establish that these interfaces are broadly composable or permissioned in a general way, so the signal is about product controls becoming more explicit, not proof of a fully agent-ready boundary.

Why now

The release date is November 20, 2026. The notes give no rationale for the timing of the Enterprise Admin Console login change, easel, or the runner additions, so the date alone does not establish a wider rollout trigger.

Connections to other events

New Runtime’s earlier coverage of Claude Code v2.1.318 described sign-in expiry reporting and gateway-related changes (v2.1.318 release note, November 4, 2026). The new release extends the product’s organization-managed sign-in story, but the notes do not say that the new Enterprise Admin Console flow replaces or changes the earlier expiry reporting.

Why it matters

For a team deploying Claude Code across an organization, the login path matters because settings and provider or gateway selection can follow the organization account rather than being treated as an individual setup detail. The release makes that behavior explicit after claude auth login, though teams still need to verify which settings apply in their own environment.

Easel could reduce friction when a user wants a visual representation of MCP metrics, while the runner metrics give operators two defined points to monitor between work order and session handoff. Those measurements could help identify slow queueing or startup stages, but the notes provide no baseline or performance claim. The --start-dir option is similarly operationally useful where work should begin in a repository subfolder without changing how settings and hooks load.

What happens next

  • Teams using Enterprise Admin Console sign-in will document which organization settings and provider choices take effect after CLI login. (6 months, 65%). Watch for: Anthropic documenting the applied settings and supported organization configurations.

Limits and uncertainty

Anthropic’s release notes do not enumerate the Enterprise Admin Console settings applied at login, or specify differences between supported cloud providers and gateways. They describe easel’s availability and charting triggers, but not its supported chart types, configuration beyond disabling automatic charts, or whether it handles arbitrary MCP output. The runner metrics define timing endpoints but include no observed values, service-level targets, or evidence that the new controls improve throughput. No independent performance evaluation accompanies these release notes.

Skill to build now

If you operate self-hosted runners, check whether your jobs should begin in a repository subfolder and test --start-dir or SELF_HOSTED_RUNNER_START_DIR in a non-production run. Then inspect the two new /metrics timings across a representative set of jobs: separate time to registration from time to first session handoff before deciding where to investigate delays.

Reader takeaway

The practical core of v2.1.330 is clearer organization-managed sign-in plus more observable and configurable self-hosted runner startup. Easel and the 30-day loop window add user-facing capability, but teams should validate the enterprise settings and runner timing behavior in their own setup before treating the release as a change to operating guarantees.

Evidence / sources

  1. 1github.com/anthropics/claude-code/releases/tag/v2.1.330
All Field Notes

Next reads

Read next

Open the graph
  1. 01topicClaudeExplore the claude topic hub.
  2. 02topicClaude CodeExplore the claude-code topic hub.
  3. 03topicAnthropicExplore the anthropic topic hub.
  4. 04archiveField NotesOpen the latest editorial analysis.
  5. 05sourcesSourcesSee the original sources this archive cites.
Who read this page?Machine requests, hidden until opened

Loading the privacy-safe route aggregate...

Open the JSON contract