← Toutes les mises à jour

SUIVI QUOTIDIEN DES VERSIONS

Claude Code

Anthropic’s terminal-native coding agent and developer workflow tool.

Les explications ci-dessous sont en anglais (aucune version française n’est encore produite).

GitHub releasesDernière vérification : Sep 25, 2026Actualisé chaque jour
v2.1.282Dernière

v2.1.282

Explication (en anglais)

This update enhances terminal UX with configurable prose width and improves operational resilience by fixing compaction retries and handling Postgres failovers via readiness grace periods. It also resolves critical API errors related to web search decryption and extended thinking loss, stabilizing session continuity for developers.

Points forts officiels

  • Added a `maxProseWidth` setting that caps the width of Claude's prose in wide terminals while tables and code blocks keep the full width
  • Added a startup notice, and `/status` and `claude doctor` entries, listing telemetry variables in a project's settings files that were ignored or that turned telemetry off
  • Added the `allowClaudeInChromeWithManagedMcp` managed setting to let `claude --chrome` run alongside an exclusive `managed-mcp.json`; the error shown when Chrome is blocked now names it
  • Added `store.readiness_grace_seconds` to the Claude apps gateway so `/readyz` can stay ready through a short Postgres outage such as a database failover
Voir la mise à jour d'origine →
v2.1.281

v2.1.281

Explication (en anglais)

This update enhances the Claude apps gateway with AWS IAM role assumption, Bedrock guardrails, and expanded desktop policy controls, while improving MCP server validation and fixing critical session stability issues. These changes provide operators with stricter security configurations and better telemetry, while developers benefit from more robust error handling and UI refinements.

Points forts officiels

  • Added Claude apps gateway support for newer Claude Desktop keys in `desktop` policy blocks, including `blockReadsOutsideWorkingDirectories` and `disableBypassPermissionsMode`
  • Added `assume_role` on Claude apps gateway Bedrock upstreams: the gateway calls Bedrock as an IAM role it assumes through STS, in another AWS account if needed, optionally one session per developer
  • Added `guardrail: {id, version}` on Claude apps gateway Bedrock upstreams to apply an Amazon Bedrock guardrail to every request sent through them (set it on all Bedrock upstreams or none)
  • Added `telemetry.resource_attributes` to the Claude apps gateway config, to put fixed labels on the telemetry of Claude Desktop and `/login` sessions
Voir la mise à jour d'origine →
v2.1.280

v2.1.280

Explication (en anglais)

This update introduces Claude Opus 5.5 as the default model with 1M context and new pricing, while adding environment variable support to override MCP description length limits. It also enhances observability via OpenTelemetry metrics and resolves critical auto-mode retry loops and symlink validation bugs.

Points forts officiels

  • Added Claude Opus 5.5 (`claude-opus-5-5`), now the default Opus model — 1M context, $4/$20 per Mtok with $0.20/Mtok cache reads
  • Added mouse support to more lists in fullscreen mode: the wheel scrolls the `/skills` list, and a skill's state options in `/plugin` can be clicked
  • Added `CLAUDE_CODE_MAX_MCP_DESCRIPTION_LENGTH` to change the 2,048-character cap on MCP tool descriptions and server instructions for every MCP server in the session
  • Added hook output sizes and the number of oversized outputs saved to a file to the `hook_execution_complete` OpenTelemetry event
Voir la mise à jour d'origine →
v2.1.278

v2.1.278

Explication (en anglais)

This update shifts the auto mode classifier to run server-side by default for API, Enterprise, and gateway users, eliminating classifier overhead charges. Developers can verify the execution context via the new /status row or opt out using the CLAUDE CODE AUTO MODE=0 environment variable.

Points forts officiels

  • Added an `Auto mode server` row to `/status` showing whether this session's auto mode classifier runs on the server
Voir la mise à jour d'origine →
v2.1.277

v2.1.277

Explication (en anglais)

This update introduces AGENTS.md support as a fallback for project instructions and enhances Claude apps gateway proxy configurations with egress boundary controls and static headers. It also resolves critical stability issues, including session hangs, empty text block errors, and plugin installation failures.

Points forts officiels

  • Added AGENTS.md support: in a project with no CLAUDE.md, Claude Code reads AGENTS.md instead; change it under "Project instructions" in `/config` (not yet on Bedrock, Vertex or Foundry)
  • Added `CLAUDE_GATEWAY_PROXY_IS_EGRESS_BOUNDARY=1` for Claude apps gateways whose only egress is a forward proxy: every outbound request hands the proxy the hostname instead of resolving it locally
  • Added an optional `headers:` map on Claude apps gateway upstreams, to send static headers to a proxy you run in front of a provider
  • Added a line saying a background task's update is waiting when it finishes while a panel such as `/tasks` is open
Voir la mise à jour d'origine →
v2.1.276

v2.1.276

Explication (en anglais)

This patch resolves a 2.1.275 regression where all API requests failed with a 400 error due to an invalid input tag when using a custom ANTHROPIC_BASE_URL proxy or gateway. It restores full compatibility for developers and operators routing traffic through intermediate infrastructure.

Points forts officiels

  • Fixed every request failing with `400 … Input tag 'advisor_20260301'` when `ANTHROPIC_BASE_URL` points at a proxy or gateway (2.1.275 regression)
Voir la mise à jour d'origine →
v2.1.275

v2.1.275

Explication (en anglais)

This update enhances security and telemetry reliability by adding gateway sign-in confirmation, startup warnings for failed otelHeadersHelper, and fixes for credential leakage in logs. It also improves workflow efficiency through a send-now interrupt key, better subagent text streaming, and automatic syncing of skills and plugins from claude.ai.

Points forts officiels

  • Added the signed-in account to Claude apps gateway sign-in: when the gateway names it, you confirm it before the credential is saved, and `/status` shows it
  • Added a send-now key (ctrl+enter, or ctrl+x ctrl+s) that interrupts the current turn and sends all queued messages at once; sent and queued messages show in gray until the model receives them
  • Added a startup warning when a configured `otelHeadersHelper` fails, so sessions that silently export no telemetry are noticed
  • Added syncing of the skills and plugins enabled on your claude.ai account to terminal sessions signed in with it; opt out with `syncClaudeAiSkills: false` or `syncClaudeAiPlugins: false`
Voir la mise à jour d'origine →
v2.1.274

v2.1.274

Explication (en anglais)

This update enhances operational stability by introducing critical memory warnings, fixing MCP server connection and timeout issues, and resolving session retry loops. It also improves observability with new OpenTelemetry attributes and managed settings events for better debugging.

Points forts officiels

  • Added a visible warning when memory usage is critical, with steps to free memory or restart safely
  • Added `CLAUDE_CODE_MCP_STARTUP_WAIT_MS` to bound how long the first non-interactive turn waits for connecting MCP servers (`0` = don't wait)
  • Added `effort` attribute to the `claude_code.llm_request` OpenTelemetry trace span, matching the `api_request` event
  • Added `claude_code.managed_settings_resolved` OTel event: managed-settings sources and policy helper state; redacted settings and digests with `OTEL_LOG_MANAGED_SETTINGS=1`
Voir la mise à jour d'origine →