Cline Release Notes
Last updated: Mar 6, 2026
- March 2026
- No date parsed from source.
- First seen by Releasebot:Mar 6, 2026
3.71.0
Cline releases GPT-5.4 models for ChatGPT subscribers, adds a Notification hook for attention and completion boundaries, and fixes MCP reconnects.
Added
- Added GPT-5.4 models for ChatGPT subscription users
- Hooks: Added a
Notificationhook for attention and completion boundaries
Fixed
- Handle streamable HTTP MCP reconnects more reliably after disconnects
- March 2026
- No date parsed from source.
- First seen by Releasebot:Mar 5, 2026
3.70.0
New Cline API docs released with Getting Started and SDK examples. Hooks now expose model.provider and model.slug, and token updates occur in real time as usage chunks arrive. Reliability fixes and docs updates enhance subagent behavior and error handling.
Added
- New Cline API docs: Getting Started, Auth, Chat Completions, Models, Errors, and SDK Examples
- Hook payloads now include
model.providerandmodel.slug - Token/cost updates now happen immediately as usage chunks arrive, not after tool execution
Fixed
- Improve subagent context compaction logic
- Subagent stream retry delay increased to reduce noise from transient failures
- State serialization errors are now caught and logged instead of crashing
- Removed incorrect
max_tokensfrom OpenRouter requests
Changed
- Windows test cleanup now retries on locked files and applies per-test timeouts
- Updated hooks docs
All of your release notes in one feed
Join Releasebot and get updates from Cline and hundreds of other software products.
- March 2026
- No date parsed from source.
- First seen by Releasebot:Mar 3, 2026
3.69.0
Added
- Add
User-Agentheader to requests sent to the Cline backend - Add default auto-tag workflow for publish release flow
- Show Cline SDK docs on the Cline page
Fixed
- Retry nested git restore and prevent silent
.git_disabledleftovers in checkpoints - Prevent Chinese filename escaping in diff view
- Trigger auto-compaction on OpenRouter context overflow errors
- Restore GPT-OSS native file editing on OpenAI-compatible models
Changed
- Update Cline SDK docs
- Improve hooks support for Windows PowerShell
- February 2026
- No date parsed from source.
- First seen by Releasebot:Feb 27, 2026
3.68.0
New dynamic Cline provider fetch, improved CLI Markdown formatting, and focus indicators for extension actions signal a tighter, more polished workflow. The release fixes auth loops, file context errors, unsafe YAML loading, and token management across Claude, OpenRouter, and other providers.
Added
- Add dynamic Cline provider model fetching from Cline endpoint
- Add additional Markdown formatting in CLI
- Add focus indicator on action buttons in extension
Fixed
- Clear all OCA secrets on auth refresh failure to prevent re-auth loops
- Resolve "Could not find the file context" error in Explain Changes
- Use
JSON_SCHEMAforyaml.loadto prevent unsafe deserialization - Fetch model info from API in CLI headless auth for Cline and Vercel providers
- Generate commit message from staged changes only when staging exists
- Update stale
maxTokensvalues for Claude 3.7+ models across Anthropic, Bedrock, Vertex, and SAP AI Core - Use
model.info.maxTokensfor OpenRouter instead of hardcoded8192
Changed
- Increase timeout for a flaky test to reduce short-term test instability
- February 2026
- No date parsed from source.
- First seen by Releasebot:Feb 25, 2026
3.67.1
Added
- Added Cline SDK API interface for programmatic access to Cline features and tools, enabling integration into custom applications.
- Added Codex 5.3 model support
Fixed
- Fix OpenAI Codex by setting
storetofalse - Use
isLocatedInPath()instead of string matching for path containment checks
- February 2026
- No date parsed from source.
- First seen by Releasebot:Feb 24, 2026
3.67.0
New release adds skills support with optional modelId, file-based AgentConfigLoader, OpenAI native provider responses, websocket preconnect for lower latency, backend-fetched featured models with local fallback, a /q quit command, enterprise config details, and dynamic banner cache. Numerous fixes and changes boost stability and provider compatibility.
Added
- Add support for skills and optional modelId in subagent configuration
- Add AgentConfigLoader for file-based agent configs
- Add Responses API support for OpenAI native provider
- Preconnect websocket to reduce response latency
- Fetch featured models from backend with local fallback
- Add /q command to quit CLI
- Add MCP enterprise configuration details
- Pull Cline's recommended models from internal endpoint
- Add dynamic flag to adjust banner cache duration
Fixed
- Fix reasoning delta crash on usage-only stream chunks
- Fix OpenAI tool ID transformation restricted to native provider only
- Fix auth check for ACP mode
- Fix CLI yolo mode to not persist yolo setting to disk
- Fix inline focus-chain slider within its feature row
- Fix Gemini 3.1 Pro compatibility
- Fix Cline auth with ACP flag
Changed
- Move PR skill to .agents/skills
- SambaNova provider: update models list
- Remove changeset-converter GitHub Action and npm run changeset
- February 2026
- No date parsed from source.
- First seen by Releasebot:Feb 20, 2026
- February 2026
- No date parsed from source.
- First seen by Releasebot:Feb 19, 2026
3.65.0
Added
- Add /skills slash command to CLI for viewing and managing installed skills
Fixed
- Fix aggressive context compaction caused by accidental clicks on the context window progress bar silently setting a very low auto-condense threshold
- Fix infinite retry loop when write_to_file fails with missing content parameter.
- Fixed default claude model
- February 2026
- No date parsed from source.
- First seen by Releasebot:Feb 19, 2026
- February 2026
- No date parsed from source.
- First seen by Releasebot:Feb 17, 2026
3.63.0
Added
- added zai GLM 5 Free promo
Fixed
- Restore reasoning trace visibility in chat and improve the thinking row UX so reasoning is visible, then collapsible after completion.