# Claude Code 2.1.317 asks before web or shell calls send account email

> A new check in Claude Code 2.1.317 asks for approval before a web or shell tool call sends the account email address over the network.

- Status: official (From the company, authors or authority themselves, with a link to their own page.)
- Topic: Safety & Security
- Published: 2026-11-03T00:00:00+00:00
- Updated: 2026-11-07T11:32:03.899181+00:00
- Web page: https://ai-news.ikangai.com/story/claude-code-2-1-317-asks-before-web-or-shell-calls-send-account-email/
- Seen in: Claude Code changelog
- Change: security (Security)
- Takes effect: 2026-11-03
- Applies to: Claude Code 2.1.317

## What changed

Claude Code now pauses a web or shell tool call for approval before it can send the account email address over the network.

Claude Code 2.1.317 adds a check that pauses a web or shell tool call before it can send the account email address to the network. The changelog for that release, dated November 3, 2026, lists it under Added. The release note says the check asks before such a call sends the address.

The changelog does not say which tool calls carry the email today or how the check decides that a call would send it. It also does not say whether the prompt can be remembered for later calls. Readers cannot tell from the note how often the prompt will appear.

The same release adds a warning for scheduled work. A warning now appears in /schedule when a new cloud scheduled task gets connectors the user did not ask for. The entry says the warning includes how to remove them.

Several other entries concern control and visibility. A new option, onFailure set to "block", makes an Agent SDK hook callback that fails stop the action it guards, where previously the action would run. A notice in Claude Desktop's Code tab now reports when an organization's remote managed settings haven't loaded for a session, and another when they do. A notice also appears when auto mode's check has no answer for a subagent's tool call. The changelog says that gap used to leave nothing in the main view.

Other entries are bug fixes. One stops claude auth status from exiting before a slow token refresh is saved, which the entry says could sign the user out on the next run. Another makes claude -p exit 1 when a full disk or quota loses transcript or ~/.claude.json writes, where it previously exited 0. The error is printed, and in text and json output the exit is 1. The entry adds that full NFS exports may be missed. A fix for self-hosted runners covers a session that moves to a new runner. The entry says Claude did not see the interrupted tool call when the turn was older than CLAUDE_CODE_RESUME_INTERRUPTED_TURN_MAX_AGE_MS.

The release also adds a cost report to the Claude apps gateway admin API at GET /v1/organizations/analytics/cost. It reads past spend over a date range in total, per developer, per day or per IdP group. The report draws on a daily record of each developer's IdP groups, kept on gateways with admin: set for admin.identity_retention_days, which the cost report groups by.

## Why it matters

Teams handling sensitive code or personal data get clearer control over when developer identity leaves their machines through agent tool calls.

## What you can do

Security leads using Claude Code could brief developers on the new prompt so that approvals of email-sending calls are deliberate.

## Original source

- [Claude Code changelog - Claude Code Docs](https://code.claude.com/docs/en/changelog)
- [claude-code/CHANGELOG.md at main · anthropics/claude-code · GitHub](https://github.com/anthropics/claude-code/blob/main/CHANGELOG.md)

---

AI wrote this summary from the sources listed, and nobody reviewed it individually before publication. Check the original source before you rely on it.

Cite as: IKANGAI AI News, "Claude Code 2.1.317 asks before web or shell calls send account email", 2026-11-03, https://ai-news.ikangai.com/story/claude-code-2-1-317-asks-before-web-or-shell-calls-send-account-email/
