RepoJournal
Anthropic @anthropics
150 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 2026-05-31
stories 14
GitHub RSS Archive Reviews

© 2026 RepoJournal Home Showcase How it works Privacy

$ the-wire · showcase

CLAUDE PLUGINS API MIGRATION WINDOW CLOSES - MOVE TO OUTPUT_CONFIG NOW

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

The Messages API is sunsetting the top-level output_format field, and your structured outputs need to move to output_config.format before the transition period ends.

Claude plugins official just merged the security-guidance migration [1] that shifts structured output schema specification from the deprecated output_format field to the nested output_config.format field [2]. The old form still returns HTTP 200 on api-key and non-streaming auth, but that's a temporary window - the clock is ticking. This isn't optional polish. Simultaneously, the team shipped three consecutive fixes [3] [4] [5] for a critical Windows path-handling vulnerability in git invocations that bypassed quotePath protections on multiple git call sites. The latest patch moves core.quotePath=false to GIT_CMD globally rather than per-site, catching edge cases in rev-parse, reflog, and git show operations that the initial fix missed. On the infrastructure side, the defending-code reference harness got its first documentation pass [6], and the cargo-nix plugin switched test execution from runCommand to mkDerivation [7] for better extensibility and crate-root directory handling.

Action items

References

  1. [1] security-guidance: migrate from deprecated output_format to output_config.format (#2098) ↗ anthropics/claude-plugins-official
  2. [2] Merge pull request #2105 from anthropics/fix-2098-output-config-format ↗ anthropics/claude-plugins-official
  3. [3] Merge pull request #2101 from anthropics/fix-2099-quotepath-global ↗ anthropics/claude-plugins-official
  4. [4] security-guidance: move core.quotePath=false to GIT_CMD globally (#2099 followup) ↗ anthropics/claude-plugins-official
  5. [5] Merge pull request #2100 from anthropics/fix-2099-text-true-pythonutf8 ↗ anthropics/claude-plugins-official
  6. [6] Merge pull request #1 from anthropics/mmolash/docs-updates ↗ anthropics/defending-code-reference-harness
  7. [7] Merge pull request #3 from Kranzes/mkdrv ↗ anthropics/cargo-nix-plugin

$ ls anthropics/reviews/ # the step-back reads

the week in review Claude Code symlink TOCTOU allows arbitrary file write the month in review Sandbox and SDK holes dominate Anthropic's September
Full archive → Weekly reviews →

Quick answers

What shipped in Anthropic on May 31, 2026?
The Messages API is sunsetting the top-level output_format field, and your structured outputs need to move to output_config.format before the transition period ends. In total, 9 commits and 5 pull requests landed.
What were the notable Anthropic updates?
security-guidance: migrate from deprecated output_format to output_config.format (#2098), Merge pull request #2105 from anthropics/fix-2098-output-config-format, and Merge pull request #2101 from anthropics/fix-2099-quotepath-global.
♥

$ log @anthropics --recent

sandbox-runtime fixes a quadratic glob parse, and five projects enroll in OSS Scanner 52 shipped buffa breaks inferred collect() calls, sandbox-runtime fixes the proxy and Linux relays 120 shipped Buffa hardens codegen and descriptor validation, OSS Scanner enrolls WordPress and the AT Protocol stack 26 shipped security-guidance findings reach stderr, and sandbox ignoreViolations types match the CLI 28 shipped

$ log --all # elsewhere on the wire

Alignment & Post-training verl rejects conflicting vLLM engine_kwargs, open-instruct adds MILES OPD behind a Megatron learner Interpretability nnsight 0.8.0 cuts import time from 7.6s to 1.6s

$ ls showcases/

Safety Evals AI Red Teaming Agent Security Interpretability Alignment & Post-training

$ init your-desk

RepoJournal can write one of these for any repo you care about. Sign in with GitHub, pick your repos and sources, and your desk gets a wire like this one every morning.