Claude, the essentials — edition of November 6, 2026
Anthropic gives admins IP allowlists and remote sign-out as Claude Code adds safeguards
Anthropic's release notes add two beta administrative controls: an IP allowlist in the Claude Console and an Admin API call that signs a member out everywhere. Recent Claude Code releases and a Messages API beta add consent checks and configuration options, and the research blog posts a new Residency call.
- The Claude Console now has a beta IP allowlist. Requests from addresses not on the list are refused, which limits where a leaked API key can be used.
- Claude Enterprise administrators can use the Admin API, in beta, to sign a member out of Claude everywhere. The member stays in the organization and can sign in again.
- Claude Code 2.1.317 asks before a web or shell tool call sends the account email address to the network, and Agent SDK hook callbacks can now block an action when the callback fails.
- The Messages API has a beta called message threads, for long conversations and agent loops. It keeps the conversation on the server so requests are smaller and time to first token is lower.
- Anthropic says text from Claude Mythos 5, Opus 4.7 and Sonnet 4.6 carries its text watermark on the platforms it lists.
Narrowing where credentials and sessions can be used
Anthropic's API release notes dated November 5 describe two beta controls for organizations. The first is an IP allowlist in the Claude Console. Requests to covered services from addresses not on the list are refused, which, in Anthropic's words, limits where a leaked API key can be used. Administrators can add IPv4 and IPv6 addresses or CIDR ranges, up to 1,000 per organization. The second lets an administrator of a Claude Enterprise (claude.ai) organization sign a member out of Claude everywhere through the Admin API. Anthropic specifies that the member stays in the organization and can sign in again, so this ends sessions without removing the person. One control acts on where requests may come from, the other on a member's active sessions.
Claude Code 2.1.317 shows a similar attention to consent and failure behavior. Among its additions, it asks before a web or shell tool call sends the account email address to the network. It also warns, in /schedule, when a new cloud scheduled task gets connectors the user did not ask for, and explains how to remove them. For the Agent SDK, onFailure: "block" means a hook callback that fails blocks the action instead of letting it run.
Sources: You can now set up an IP allowlist for your organization in the Claude Console, in beta — Anthropic (release notes) · You can now sign a member of your Claude Enterprise organization out of Claude everywhere with the Admin API — Anthropic (release notes) · Claude Code 2.1.317 — Claude Code
Smaller requests and finer configuration for developers
The November 2 API release notes introduce message threads, in beta behind the message-threads-2026-08-12 header, for long conversations and agent loops on the Messages API. Anthropic says the API keeps the conversation on the server, so later requests can be smaller, and that this gives a lower time to first token. The notes present it as a way to cut the cost of resending a growing history, and they flag it as a beta.
Claude Code 2.1.318 adds three items for teams that configure the tool heavily. It records the user's answer when Claude Code asks to turn on the mods it writes in a session. It adds assume_role on mantle upstreams of the Claude apps gateway, for a role in another AWS account and for per-developer AWS cost attribution. It also adds an effort option to a mod's $.model.fork, so a quick side job can run at low effort whatever effort the session uses. The result reports the effort that was sent. Claude Code 2.1.316 adds CLAUDE_CODE_DISABLE_ULTRATHINK, which turns the ultrathink keyword off.
Sources: Long conversations and agent loops on the Messages API can now send smaller requests, with message threads (beta) — Anthropic (release notes) · Claude Code 2.1.318 — Claude Code · Claude Code 2.1.316 — Claude Code
Marking model output, and a post on provable inference
A November 3 entry in the API release notes states that text generated by Claude Mythos 5, Claude Opus 4.7 and Claude Sonnet 4.6 carries Anthropic's text watermark on the platforms listed in its page "How Claude marks AI-generated content". The entry also singles out Claude Sonnet 4.6 for requests that set temperature, top_p or top_k to a non-default value. The excerpt available to us is cut off before it says what applies in that case, so we do not describe the exception.
Anthropic also published a post titled "Progress on provable inference for large language models". Only the title was available to us, so we make no claim about its content or findings.
Sources: Text generated by Claude Mythos 5, Claude Opus 4.7, and Claude Sonnet 4.6 carries Anthropic's text watermark — Anthropic (release notes) · Progress on provable inference for large language models — Anthropic
Science with Claude, and a call for senior technical talent
Anthropic published two posts on scientific work: "AI for scientific discovery: How researchers use Claude" and "Investigating vault particles with Claude". Also listed is a 2026 paper by Athukoralage et al., "Prokaryotic homologs of the major vault protein form engineerable vault nanocapsules". It shares the vault subject matter, but we do not know how it relates to the post. We had only the titles of these items, so we do not summarize their results.
On the research side, Anthropic announced that applications are open for the Anthropic Residency, aimed at established technical leaders. The item is attributed to the Anthropic Alignment Science blog. The material provided gives no further detail on the programme's terms, duration or selection.
Sources: AI for scientific discovery: How researchers use Claude — Anthropic · Investigating vault particles with Claude — Anthropic · Athukoralage et al. 2026, Prokaryotic homologs of the major vault protein form engineerable vault nanocapsules — Anthropic · Anthropic Residency: applications open for established technical leaders — Anthropic
This edition is an original synthesis written by Claude from aggregated news — Anthropic's own sources first (release notes, status, newsroom, research, engineering), then the press, Hacker News, Reddit and GitHub, under the editorial supervision of Héra SASU. Every fact links to its article, publisher named. See the live feed →
Claude News is published by Héra SASU. Independent media, not affiliated with Anthropic.