megachangelog
Fix0.0.74

SDK v0.0.74

Multiple fixes to Claude Code provider for agentic work, tool-call JSON parsing, session management, and telemetry. Improved Hub daemon upgrade behavior and reporting of involuntary logouts, plus better token telemetry accounting.

  • Fixed the Claude Code provider being unusable for agentic work: the provider now declares its own native tools instead of receiving Cline's unbridgeable tool definitions, the session is anchored on the workspace directory instead of inheriting the host's cwd, and ~/.claude plus project settings are loaded so user-configured permission rules apply. File edits under the workspace are auto-approved; command execution stays gated by your own Claude settings
  • Fixed truncated tool-call JSON being silently "repaired" into wrong arguments — a payload with an unterminated string is now rejected rather than getting an invented terminator
  • Fixed strict providers rejecting a turn with "user message must have content" when a message's content held only empty text parts
  • Fixed a mid-turn crash on streamed tool calls with non-zero or non-contiguous indexes, hit through LiteLLM's Anthropic passthrough, by updating the AI SDK packages
  • Managed Hub daemons now upgrade directionally: when another install ships a newer Hub build, hosts attach to the newer daemon and prompt to update and restart instead of the two installs repeatedly retiring each other's daemons. Older or unordered Hubs are still retired and replaced
  • Fixed the Hub daemon logging an unhandled hub server close failed rejection and exiting non-zero whenever a client was connected at shutdown; shutdown is now clean
  • run.started is now emitted only after the target session resolves and carries the originating requestId and clientId, so multi-client hosts can correlate delivery acknowledgments
  • Token telemetry now reports disjoint per-request buckets — uncached input, cache reads, cache writes — instead of re-counting the whole cached conversation on every event, which inflated per-task sums roughly 5x on cache-heavy sessions
  • Involuntary Cline logouts (a rejected refresh token) are now reported instead of credentials being cleared silently; a transient network failure refreshing the stored session on startup no longer books as a logout
  • Per-token stream deltas are no longer mirrored into telemetry — they accounted for ~97% of all agent event volume with no analytical value

Full Changelog: sdk/sdk/v0.0.73...sdk/sdk/v0.0.74

sdkclaude-codetoolstelemetrystability

Source: original entry ↗

More from Cline

Follow Cline to get its new changes in your feed and email digest.

Improvement0.0.48

Desktop v0.0.48

Added support for GPT-6 models on OpenAI and Azure OpenAI endpoints by fixing the max_tokens parameter issue. Updated default models across multiple providers and refreshed the model catalog with new CommandCode and Reflection providers.

modelsapiopenaiazureproviders
Improvement3.0.73

CLI v3.0.73

GPT-6 models now work with OpenAI and Azure OpenAI endpoints after fixing a token parameter issue. Claude Sonnet 5.5 is restored as the default model when none is selected, and the model catalog has been refreshed with new providers and updated default models across multiple services.

climodelsopenaiazureproviders
Improvement0.0.95

SDK v0.0.95

OpenAI-compatible requests for GPT-6 and later models now use max_completion_tokens instead of max_tokens to prevent rejection by OpenAI and Azure. The hub daemon now tracks client launcher information as OTel resource attributes, the model catalog expanded from 6,952 to 7,041 models with new providers, and default models were updated across multiple providers with Claude Sonnet 5.5 restored as the primary recommended option.

sdkapimodelstelemetryopenai-compatible
See all Cline changes →