🧭 Anthropic Expands Claude's Enterprise Security Governance to 28 Integrations

Anthropic announced that the Claude Compliance API now connects to 28 security and compliance platforms — a major expansion from the five launch partners announced on May 21. The move positions Claude Enterprise as a governable, IT-auditable tool within corporate environments, subject to the same monitoring workflows that IT and security teams already apply to SaaS tooling.

The full integration roster

The 28 platforms span every major security category: endpoint and cloud security (CrowdStrike, Palo Alto Networks, Fortinet, Trellix, Tenable), identity and access (Okta, SailPoint), cloud access security and SASE (Zscaler, Netskope, Cloudflare, Forcepoint), data security and DLP (Microsoft Purview, Mimecast, Proofpoint, Smarsh, Theta Lake, Varonis, Cyera), developer and code security (Snyk), backup and resilience (Rubrik), SIEM and observability (Sumo Logic, Datadog, Cribl, IBM QRadar), legal and eDiscovery (Relativity), and MDR/MSSP (ReliaQuest, Geordie AI, Wiz).

What the Compliance API exposes

The Claude Compliance API gives connected platforms programmatic access to two streams:

• Conversation content — chat messages, uploaded files, and Claude Projects content from Claude Enterprise users
• Activity event logs — user logins, admin actions, configuration changes, and usage events from both Claude Enterprise and the Claude Platform

🧭 Claude Code v2.1.152: /code-review --fix, disallowed-tools in Skills & /reload-skills

Claude Code v2.1.152, released today, ships four changes that make the code-review workflow fully automated, tighten skill security, and let you reload skills without restarting a session.

What's new

• /code-review --fix — until now, /code-review (formerly /simplify) produced suggestions you had to apply manually. The new --fix flag instructs Claude Code to apply every finding — reuse, simplification, and efficiency improvements — directly to the working tree. You review the diff; Claude does the edit pass.
• disallowed-tools in skill frontmatter — skills and slash commands can now declare a list of tools they must not use, enforced at invocation time. This lets skill authors ship safe-by-default automation: for example, a read-only audit skill that cannot accidentally call Edit or Bash.
• /reload-skills command — reload all skills from disk without leaving your current session. Previously, picking up a skill change required ending and restarting the session, losing in-flight context.
• SessionStart hooks can return reloadSkills: true — gives hook scripts a programmatic way to force a skill reload at session initialisation, useful for CI/CD pipelines that pull the latest skill versions before starting work.
• MessageDisplay hook event — a new hook that fires whenever Claude Code renders a message to the user, giving hook authors a way to intercept or augment displayed output (logging, audit, UI integrations).

# Skill frontmatter example — read-only audit skill
---
name: audit-deps
description: Audit package dependencies for security issues (read-only)
disallowed-tools:
  - Edit
  - Write
  - Bash
---

🧭 Anthropic Appoints KiYoung Choi as Korea Representative Director — Seoul Office Opening

Anthropic has appointed KiYoung Choi as Representative Director of Korea, the company's senior local leadership role, ahead of opening a Seoul office. The appointment signals Anthropic's commitment to building dedicated regional presence in South Korea — one of Asia-Pacific's largest enterprise software markets and a country with significant AI investment from both government and conglomerates including Samsung, LG, and SK Telecom.

Why Korea, why now

• South Korea has an aggressive national AI strategy with multi-billion-dollar public investment in AI infrastructure announced in 2025 and 2026
• Korean enterprises are among the fastest global adopters of enterprise AI tooling — several major chaebols were early Claude API customers
• The Seoul office follows Anthropic's Tokyo office (opened late 2025) as part of a deliberate Asia-Pacific expansion, complementing existing hubs in San Francisco, London, and Dublin
• South Korea has a robust domestic AI regulation framework under development; a local Representative Director provides the regulatory engagement capability required