Source Intelligence

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 v2.0.29 Home All releases olderv2.0.28 v2.0.30newer
Claude Code v2.0.29

Enhanced Trust Validation Logic

What: The workspace trust check now respects the global bypass permissions setting more efficiently. When users have accepted bypass permissions mode (via --dangerously-skip-permissions or --allow-dangerously-skip-permissions flags), the trust validation immediately returns true without checking individual directory trust settings.

How it works:

# Enable bypass permissions mode (shows one-time warning dialog)
claude --dangerously-skip-permissions

# Or make it available as an option
claude --allow-dangerously-skip-permissions

Details:

  • Previously, bypass permissions mode existed but trust checks were independent operations
  • Now, accepting bypass permissions mode acts as a global trust override for all workspace trust dialogs
  • Eliminates redundant directory traversal when bypass mode is already active
  • The bypass permissions feature requires explicit user consent via an interactive warning dialog
  • Designed for sandboxed/containerized environments with restricted internet access
  • Evidence: Enhanced check in rJ() at line 468251 in v2.0.29

Technical changes:

  • Added early-return check for bypassPermissionsModeAccepted flag before performing directory-specific trust validation
  • Reduces unnecessary computation when global bypass is enabled
  • No change to the security model—only optimization of the existing trust system

See this entry in the whole of v2.0.29 →