Source Intelligence
Follow Discord

DisclaimerUnofficial, and not affiliated with Anthropic. Nearly all of this is read straight out of what ships: npm bundles, captured prompts, published docs. Anthropic's own notes go in verbatim, marked as theirs. The rest is my reading, and every entry carries the strings behind it. If one looks wrong, vote it down and say why.

← All of v2.1.86 Home All releases ←olderv2.1.85 v2.1.87newer→
Claude Code v2.1.86

  • Directory removal is now more robust — uses a single rmdir call and handles ENOTDIR/ENOENT/ENOTEMPTY errors gracefully instead of manually checking if a directory is empty first (search for "Could not remove directory at")
  • Lock acquisition failure logging simplified — always wraps errors with a cause chain instead of conditionally stringifying (search for "NON-FATAL: Lock acquisition failed")
  • Skill usage tracking now debounces writes with a minimum interval, preventing rapid config updates when a skill is invoked repeatedly (search for "skillUsage" in bI8())
  • SwarmPermissionPoller now validates and drops malformed permission update entries with a warning instead of crashing (search for "[SwarmPermissionPoller] Dropping malformed permissionUpdate")
  • Advisor feedback message wording corrected: "will apply the feedback" (was: "will now apply the feedback.") (search for "Advisor has reviewed the conversation and will apply the feedback")

See this entry in the whole of v2.1.86 →