Claude Code v2.1.185
Released Jun 20, 2026 (2 months ago).
183 issues reference this build
29 still open
154 resolved (84%)
"References this build" means the issue body mentions running 2.1.185 (from the
bug-report template). It reflects report volume, not a proven defect count — an issue may mention a
version without being caused by it. First-party data from the anthropics/claude-code tracker.
Top issues reported on v2.1.185
Ranked by community engagement (reactions and comments).
- [Bug] Anthropic API Error: Service Unavailable
- [Bug] Anthropic API Error: 529 Overloaded
- [BUG] Model selector missing from Claude Code routine editor (Pro account)
- [BUG] macOS 26: /login completes ("Login successful") but token never written to Keychain — browser does not auto-open, device code flow used instead
- [Feature Request] Multiple account support with email-based account switching
- [Bug] Model silently switches from 1M to non-1M Opus mid-session, causing unrecoverable API Error
- CLI /model picker doesn't list Sonnet 5 as a preset; shows custom model instead
- [BUG] Remote-control session loses context after restart while web UI continues displaying conversation as continuous
- Windows: .local\bin\claude.exe installs as plain Bun v1.4.0 instead of Claude Code
- [BUG] Claude Max weekly usage depleted unexpectedly fast for Claude Code subagent-heavy Gmail MCP sessions
- [BUG] Skill invocation prompt is blocking the output.
- [BUG] Native macOS/Linux builds (>=v2.1.117) removed Glob/Grep tools, but model guidance still points to them — Claude reaches for tools that no longer exist
- Assistant prose replying to a mid-turn user message is dropped from the transcript view (text confirmed present in session JSONL)
- [Bug] Dispatch shows "Your computer" for all devices instead of actual device names
- [Bug] Frequent Anthropic API 500 errors and performance degradation
- [Bug] Excessive cache writes causing unexpectedly high token usage and costs
- [BUG] OAuth login flow gives no feedback when manual code paste is required — "if prompted" misleads users
- `claude --print` fails with 401 when no interactive session is open (regression in 2.1.185)
- [BUG] File suggester fails to dismiss after space/newline, causing ENTER to overwrite final word instead of submitting prompt
- Desktop app: session in Recents keeps showing a blinking/flashing status dot that won't clear
- [BUG] Subagents stall waiting for MCP servers to be approved when merged into worktrees
- You've hit your session limit · resets 7:30pm (Europe/Berlin)
- [Bug] Anthropic API Error: Insufficient Context
- [Bug] Advisor tool incorrectly reports network errors
- [BUG] Sonnet (default) - Usage credits required for 1M context
- [BUG] CLI: OAuth login fails with "Invalid OAuth Request — Missing redirect_uri parameter" — Pro plan, Windows 11
- [BUG] Scheduler catch-up storm on restart: re-fires daily tasks and runs enabled:false tasks (ghost fires, lastRunAt not updated)
- [BUG] claude blocks on startup due to ignoring proxy settings
- [BUG] HTTP 407 error for pretooluse http hook with authentication in sandbox
- [BUG] Model fabricates entire conversation turns (fake user messages + fake tool results) after an interrupted tool call under transmission latency
Changelog
What's changed
- The stream-stall hint now reads "Waiting for API response · will retry in …" instead of "No response from API · Retrying in …", and triggers after 20s of silence instead of 10s