186 wires and counting

$ follow Anthropic

Keep up with Anthropic in about 3 minutes: what actually shipped — the commits, pull requests, releases, and security advisories that matter.

or

fair warning: these emails are deeply technical. diffs, version numbers, CVEs, benchmark deltas. if that's not your idea of a good read, this isn't your newsletter.

Folds into your digest — weekly by default, monthly if you prefer. Unsubscribe in one click.

$ status

wire 2026-11-13
stories 56

© 2026 RepoJournal Home Showcase How it works Privacy

$ the-wire · showcase

Claude Code 2.1.323 adds experimental Code Mode and plugin version gating

By RepoJournal · Filed · About Anthropic · Composed from the cited sources · methodology

Claude Code shipped an opt-in mode where Claude writes JavaScript to drive its own tools, alongside plugin version-range enforcement and a security-guidance plugin release that narrows where its git hooks run.

v2.1.323 anthropics/claude-code

by claude-code-publisher[bot]

Code Mode is experimental and opt-in: Claude writes JavaScript that calls its tools, toggled per user or off per organization via the docs. The same release adds an engines["claude-code"] field to plugin.json, so plugins outside their stated version range are not installed, updated or loaded, and install and update resolve to the newest version that fits; claude plugin validate now also type ch...

security-guidance: only run git in ordinary checkouts anthropics/claude-plugins-official

by Claude

The security-guidance plugin now treats a folder as a repository only when it has its own .git entry or the caller named it through GIT_DIR, and sets safe.bareRepository=explicit on the hook's git commands. Linked worktrees and ordinary checkouts behave as before, while bare repositories and stray parent .git directories are no longer picked up.

dpdk: clarify mbuf trust model in threat model anthropics/oss-scanner

by shemminger

Scanner findings for DPDK are being reclassified, not fixed in code: the threat model gains an "Mbuf contract" stating that received packet data is never trusted and that Rx side must validate strictly, while Tx burst and Tx prepare trust the mbuf. Stephen Hemminger wrote that earlier scans reported Tx drivers for mbufs the Rx driver had produced incorrectly; vhost guests and memif clients coun...

feat: enroll Actix Web anthropics/oss-scanner

by robjtede

Actix Web enrolled itself and the workspace crates named in its threat_model.md, with actix-test and actix-http-test declared out of scan scope. The threat model covers server and client protocol parsing, request limits and compression, and the build pins rust:1-trixie with an offline cargo test --locked --workspace --all-features run.

Enrol s3fs-fuse anthropics/oss-scanner

by gaul

s3fs-fuse, Matomo and Actix Web all filed enrolment PRs in the same window, each from a core maintainer with project.yaml passing tools/validate.py locally. security-guidance 2.0.20 was also propagated into knowledge-work-plugins and sandbox-runtime cut v0.0.83; the rest of the day was changelog and mods automation.

Quick answers

What shipped in Anthropic on November 13, 2026?
Claude Code shipped an opt-in mode where Claude writes JavaScript to drive its own tools, alongside plugin version-range enforcement and a security-guidance plugin release that narrows where its git hooks run. In total, 33 commits, 21 pull requests, and 2 releases landed.
Who contributed to Anthropic on November 13, 2026?
9 developers shipped this update, including GitHub Actions, claude-code-publisher[bot], Stephen Hemminger, gaul, sgiehl, robjtede, Mohamed Hegazy, and Claude, and 1 more.
What were the notable Anthropic updates?
v2.1.323, security-guidance: only run git in ordinary checkouts, and dpdk: clarify mbuf trust model in threat model.