## What's Changed
### Features
* feat(members): filter project and organization members by role by @hassiebp in https://github.com/langfuse/langfuse/pull/17810
* feat(scores): Allow creating scores from annotation sidebar by @bezbac in https://github.com/langfuse/langfuse/pull/17815
* feat(feature-flags): introduce internal view mode and related components by @marliessophie in https://github.com/langfuse/langfuse/pull/17816
* feat(evals): route TypeSafe decision-model connections through OpenRouter or Vercel AI Gateway by @hassiebp in https://github.com/langfuse/langfuse/pull/17798
### Fixes
* fix(scim): scope Users/{id} reads and mutations to the caller's org by @Steffen911 in https://github.com/langfuse/langfuse/pull/17828
* fix(web): Formatted / Raw toggle on the observation Attributes tab by @nmtrang29 in https://github.com/langfuse/langfuse/pull/17712
* fix(web): type scale carries line-height, whole-pixel sizes by @nmtrang29 in https://github.com/langfuse/langfus
## What's Changed
* refactor(web): reroute leftover feature deep imports through index.ts by @nkabardin in https://github.com/langfuse/langfuse/pull/17749
* fix(web): trace page panels fill the window width by @nmtrang29 in https://github.com/langfuse/langfuse/pull/17714
* refactor(web): reroute remaining feature deep imports through index.ts by @nkabardin in https://github.com/langfuse/langfuse/pull/17753
* feat: trace messages internal preview by @marliessophie in https://github.com/langfuse/langfuse/pull/17648
* fix(web): antialiased body text by @nmtrang29 in https://github.com/langfuse/langfuse/pull/17756
* refactor(web): reroute leftover feature deep imports through index.ts by @nkabardin in https://github.com/langfuse/langfuse/pull/17760
* fix(web): remove unused surveys.create tRPC route by @Steffen911 in https://github.com/langfuse/langfuse/pull/17768
* chore: bump clickhouse test version to 26.9 by @Steffen911 in https://github.com/langfuse/langfuse/pull/17769
* refa
## What's Changed
- Structured outputs on thinking models now apply in a single pass, making them faster and more reliable.
- Fixed intermittent "model not found" errors with a large local library
- Fixed the macOS app becoming unresponsive when checking if ChatGPT or Codex is running.
- Qwen 3.8 prompt processing is faster on Apple Silicon.
- Gemma 4 on Apple Silicon now picks the best image resolution per image, keeping more detail in high-resolution images.
- Updated llama.cpp, MLX, and XGrammar.
**Full Changelog**: https://github.com/ollama/ollama/compare/v0.34.3...v0.34.4
## Features
- feat(jev): add TypeSafe evaluation models to Agent (#8103)
- feat(models): add GPT-6 Sol and Luna (#8147)
- feat(providers): add Claude Opus 5.5 (#8141)
## Improvements
- improvement(settings): consolidate access requests in settings (#8154)
- refactor(ui): narrow local style forwarding and preview geometry (#8144)
- improvement(ui): share expandable home sections (#8067)
- improvement(ui): share public chat composer actions (#8075)
- improvement(ui): use chips for table job actions (#8082)
- refactor(ui): make measurement styles explicit and reuse preview white (#8142)
## Bug Fixes
- fix(knowledge): unschedule a connector whose credential the source rejected and prompt to reconnect (#8158)
- fix(billing): keep archived selected workspaces in the invitation applied-state check (#8159)
- fix(oauth): rotate a refresh chain only from the token it started from and size the refresh lease for every provider (#8152)
- fix(billing): bound every ledger aggregate at the databa
## 1.22.0 (2026-09-22)
<!-- Release notes generated using configuration in .github/release.yml at main -->
## What's Changed
### Features
* feat: test remote MCP servers on cloud backends via the app server by @hieptl in https://github.com/OpenHands/OpenHands/pull/17276
* feat: open Git Sync to org admins on cloud backends by @hieptl in https://github.com/OpenHands/OpenHands/pull/17216
* feat(i18n): translate recently added Turkish UI strings by @alkank in https://github.com/OpenHands/OpenHands/pull/17581
### Bug Fixes
* fix: show the add automation menu to org members by @hieptl in https://github.com/OpenHands/OpenHands/pull/17526
* fix(canvas): keep the base path on links that leave the app by @tofarr in https://github.com/OpenHands/OpenHands/pull/17617
* fix: preserve sibling MCP servers when saving to a cloud backend by @hieptl in https://github.com/OpenHands/OpenHands/pull/17467
* fix: link home automation rows to the automation view by @hieptl in https://github.com/OpenHands/Op
## 1.21.0 (2026-09-22)
<!-- Release notes generated using configuration in .github/release.yml at main -->
## What's Changed
### Features
* feat: request Docker execution workspaces from Docker servers by @neubig in https://github.com/OpenHands/OpenHands/pull/17177
### Bug Fixes
* fix: prevent stale conversation history merges by @adbcodes in https://github.com/OpenHands/OpenHands/pull/16913
* fix: query the Forgejo host for Forgejo PR and issue lists by @Harsh23Kashyap in https://github.com/OpenHands/OpenHands/pull/17486
* fix(apps): accept GitHub tree URLs as app source by @VascoSch92 in https://github.com/OpenHands/OpenHands/pull/17504
* fix: Disable PostHog autocapture by @malhotra5 in https://github.com/OpenHands/OpenHands/pull/17502
* fix(static-server): serve current assets after rebuilding Canvas by @neubig in https://github.com/OpenHands/OpenHands/pull/17402
* fix: show non-resumable local conversations as archived by @neubig in https://github.com/OpenHands/OpenHands/pull/1
## What's Changed
`GET /api/show` now advertises each model's thinking controls and default:
Available in the CLI with:
```
ollama show gemma4
```
```
thinking
levels false, true
default true
```
Available in the API with:
```sh
curl http://localhost:11434/api/show -d '{"model": "glm-5.3-flash:cloud"}'
```
```json
{
"thinking": {
"values": ["low", "high", "max"],
"default": "max"
}
}
```
Also available on ollama.com directly for cloud models.
* **Nemotron H** vision models are now supported on Apple Silicon with MLX
* Ollama's macOS app will now no longer reopen windows you've closed when activating it
* Fix for model pulls from HuggingFace
**Full Changelog**: https://github.com/ollama/ollama/compare/v0.34.2...v0.34.3
## 1.20.0 (2026-09-17)
<!-- Release notes generated using configuration in .github/release.yml at main -->
## What's Changed
### Features
* feat(agent-profiles): select secrets available to a profile by @simonrosenberg in https://github.com/OpenHands/OpenHands/pull/17237
* feat: forward Docker conversation runtime settings by @neubig in https://github.com/OpenHands/OpenHands/pull/17462
* feat(automations): select saved agent profiles by @neubig in https://github.com/OpenHands/OpenHands/pull/17396
### Maintenance
* test: isolate Mock-LLM profiles from ambient secrets by @neubig in https://github.com/OpenHands/OpenHands/pull/17482
* chore: bump released agent runtime dependencies by @neubig in https://github.com/OpenHands/OpenHands/pull/17489
**Full Changelog**: https://github.com/OpenHands/OpenHands/compare/v1.19.0...v1.20.0
---
This PR was generated with [Release Please](https://github.com/googleapis/release-please). See [documentation](https://github.com/googleapis/release-please
## What's Changed
### Features
- Support aliases as connection identifiers
- Record reasons for deployment creation failures
- Collect human feedback and pause events in tracing
- Add `llm_overlay` context variable to route agent roles to models
- Carry `task_prompt` and output in agent execution payloads
- Validate platform integrations during crew setup
- Add platform tools to JSON crew wizard
- Expose CrewAI Platform application catalog
- Add OpenRouter as a supported embedding provider
### Bug Fixes
- Accept CRLF in inline skill definitions
- Load text file URLs through the safe fetcher
- Close SQLite connections in kickoff task outputs storage
- Carry `from_cache` on the native tool path's `ToolUsageFinishedEvent`
- Raise `ValueError` instead of bare raise in uploader
- Support non-primitive types in SQLiteFlowPersistence
- Improve coding agents instructions
- Read JSON checkpoints as UTF-8
- Overwrite stale `poetry.lock` backup on Windows
- Key streamed tool calls by wire index
## 1.19.0 (2026-09-16)
<!-- Release notes generated using configuration in .github/release.yml at main -->
## What's Changed
### Features
* feat: scope an agent profile to specific MCP servers by @simonrosenberg in https://github.com/OpenHands/OpenHands/pull/17289
* feat: add GPT-6 Astra model support by @neubig in https://github.com/OpenHands/OpenHands/pull/17164
* feat: track authenticated Canvas arrivals by @neubig in https://github.com/OpenHands/OpenHands/pull/17083
* feat: surface automation disablement reason in the UI by @malhotra5 in https://github.com/OpenHands/OpenHands/pull/17425
### Bug Fixes
* fix: recognize Azure DevOps SSH remotes by @Chocolatine75 in https://github.com/OpenHands/OpenHands/pull/17375
* fix(static-server): HTML-escape injected runtime config and set Cache-Control: no-store on credential injection by @Tyagiquamar in https://github.com/OpenHands/OpenHands/pull/17175
* fix: avoid redundant automation SDK version refetches by @workvahedi in https://github.c
## What's Changed
- Added first-run setup when running `ollama`, with options to sign in or continue locally. Setup completion is shared with the desktop app on macOS and Windows.
- Added `ollama://apps` to open the desktop app’s Apps page directly on macOS and Windows.
- Fixed excessive memory growth during long generations with MLX speculative decoding.
- Updated llama.cpp.
**Full Changelog**: https://github.com/ollama/ollama/compare/v0.34.1...v0.34.2
## What's Changed
* MLX safetensors `ollama create` no longer experimental. GGUF model creation now requires using llama.cpp tooling for safetensor conversion and quantization.
* Improved MLX memory handling on Apple Silicon
* Runaway repeat token detection now requires 100 repeat tokens for reduced false positives (e.g. OCR)
* `/api/tags` is much faster on large model libraries (3.1 s → 294 ms cold in testing), and model capabilities are now reported consistently.
* Deprecated `typical_p`: it can no longer be set when creating new models, existing GGUF models retain support.
* MLX and llama.cpp updates
**Full Changelog**: https://github.com/ollama/ollama/compare/v0.34.0...v0.34.1-rc1
* feat(linear): filter Updated Issue trigger by changed fields (#15435) @AdamSelene, @claude, @OdaiAhmed99
## 🔌 Pieces
* fix(pieces): declare remaining runtime deps for datadog and scrapeless (#15519) @kishanprmr
* fix(pieces): hide body input for GET/HEAD requests (#15398) @kishanprmr
* feat(gmail): add AI actions for the gmail.modify scope (label, trash, archive) (#15512) @kishanprmr
* feat(google-drive): Optimize google drive list files (#15461) @danielpoonwj, @kishanprmr
* fix(bubble): Get Thing returns the single item instead of the whole list (#15460) @kishanprmr
* feat(agents): a flow step can run a saved agent (#14934) @hazemadelkhalel
* fix(convertkit): send api_secret in the query string instead of the GET body (#15451) @ibrahim-abuznaid, @claude, @MrChaker
* fix(pieces): keep the deprecated flag when a published piece is registered on cloud (#15466) @ibrahim-abuznaid, @claude
* fix(kimai): activity dropdown ignores project, migrate auth to Bearer token (#15459) @kishanprm
## 1.18.0 (2026-09-11)
<!-- Release notes generated using configuration in .github/release.yml at main -->
## What's Changed
### Features
* feat: click an image attachment to open it full size by @palrohitg in https://github.com/OpenHands/OpenHands/pull/17179
* feat: only the creator can turn an automation back on by @hieptl in https://github.com/OpenHands/OpenHands/pull/17190
* feat: allow editing automations on cloud backends by @hieptl in https://github.com/OpenHands/OpenHands/pull/17191
* feat: show which identity an automation runs as by @hieptl in https://github.com/OpenHands/OpenHands/pull/17138
* feat: use DB-driven free/default model flags in Canvas by @juanmichelini in https://github.com/OpenHands/OpenHands/pull/16685
### Bug Fixes
* fix: pass the active org from the gear icon to cloud settings by @hieptl in https://github.com/OpenHands/OpenHands/pull/17256
* fix: hide unlisted settings pages from the command menu when locked to a Cloud host by @hieptl in https://github.com
> [!WARNING]
> **Self-hosted deployments using the bundled Weaviate must complete a manual, staged upgrade before starting 1.17.1.** The bundled Weaviate server moves from `1.27.0` to `1.39.2` — 12 minor versions — and skipping minors **is unsupported**. Pulling and restarting can silently and permanently break vector search. See Upgrade Guide below. Fresh deployments, external Weaviate, and other vector stores are unaffected.
### New Features
**Dataset-Scoped Knowledge Base API Keys**
- Knowledge base service-API keys were scoped to the whole workspace: one key could read and write every knowledge base in the tenant, so giving an integration access to a single knowledge base meant giving it access to all of them. Keys can now be bound to specific knowledge bases from the knowledge base's API access panel, with a workspace / this-knowledge-base scope selector at creation time. A bound key returns `403` for any other dataset — both documents and retrieval — and for endpoints tha
## What's Changed
### Features
- Add telemetry to track checkpoint runtime and CLI usage.
### Bug Fixes
- Fix gateway errors reported inside an HTTP 200 response.
- Keep deploy push on the AMP create source.
- Report scrape failures instead of raising IndexError in the Oxylabs integration.
- Route all DashScope models through the native provider.
- Persist JSON checkpoints as UTF-8.
- Drop stray $ in f-string search URLs for BrightData.
- Support list-form type arrays in JSON schema conversion.
- Correct gpt-4o-mini context window from 200000 to 128000.
- Preserve streaming tool call arguments at contentBlockStop.
- Make pre-commit hooks portable on Windows.
### Documentation
- Clarify that tracing is managed separately from telemetry.
- Fix parallel search reference link.
- Remove stale parameters from handle_llm_stream_chunk docstring.
- Fix streaming output docstring examples.
- Use organization UUIDs in the skill install reference.
## Contributors
@Alphaxiaoteng, @DrewWhittleN
## 1.17.0 (2026-09-09)
<!-- Release notes generated using configuration in .github/release.yml at main -->
## What's Changed
### Features
* feat: show manifest-declared value statements on dashboard cards and rows by @hieptl in https://github.com/OpenHands/OpenHands/pull/16921
* feat: Make automation run UI task-outcome aware by @malhotra5 in https://github.com/OpenHands/OpenHands/pull/16937
* feat: enable LLM provider connections on cloud by @juanmichelini in https://github.com/OpenHands/OpenHands/pull/16757
* feat: allow editing a custom cron expression from the edit-automation modal by @VascoSch92 in https://github.com/OpenHands/OpenHands/pull/16964
* feat: unified cloud settings entrypoint for Agent Canvas by @jpelletier1 in https://github.com/OpenHands/OpenHands/pull/16883
* feat: add local Agent Canvas Planner support by @VascoSch92 in https://github.com/OpenHands/OpenHands/pull/16178
* feat(conversation-panel): add conversation tags and filtering by @georgeglarson in https://g
Hotfix on top of 0.88.4. Combines 0.88.4-hotfix.1 and 0.88.4-hotfix.2.
- fix(worker): disable piece cache prewarm
- Revert "fix(connections): stop oauth2 connections failing to save when the code has special characters" (#14926)