## What's Changed
### Features
* feat(llm): add organization-level connections by @wochinge in https://github.com/langfuse/langfuse/pull/18450
* feat(billing): record billable units in an append-only ClickHouse ledger by @Steffen911 in https://github.com/langfuse/langfuse/pull/18722
* feat(skills): add CLI import instructions by @tobi12345 in https://github.com/langfuse/langfuse/pull/18794
* feat(evals): restyle the evaluator gallery with score previews by @nmtrang29 in https://github.com/langfuse/langfuse/pull/18709
* feat(topics): show facet descriptions in Configure Topics by @LysanderKie in https://github.com/langfuse/langfuse/pull/18753
* feat(topics): assign automatic summaries to the serving map and start first discovery by @LysanderKie in https://github.com/langfuse/langfuse/pull/
* feat(api): optionally include skill file contents by @tobi12345 in https://github.com/langfuse/langfuse/pull/18826
* feat(tracing): show prompt cache hit rate by @LysanderKie in https:/
## Features
- feat(project-files): add public sharing and protected delivery (#9019)
- feat(project-files): add cross-owner copy and bulk transfers (#9018)
- feat(project-files): add rendered delivery and indexed search (#9017)
## Improvements
- improvement(api): publish an RFC 9727 API catalog and stop padding tables in docs markdown (#9066)
## Bug Fixes
- fix(chat): preserve panel resize gestures (#9065)
- fix(mothership): take over runs whose controller died and re-run their interrupted tool (#9059)
- fix(files): restore workspace read-path behavior after rendered delivery (#9064)
- fix(pii): load models only in workers and recycle them before memory runs out (#9060)
- fix(project-files): requeue rendered-artifact search after commit and trim Project read transactions (#9063)
- fix(contracts): drop unused project file search query exports (#9062)
## Other Changes
- chore(ci): tag production promotions with release-<sha> for rollback retention (#9061)
## Contributors
- @mzxc
## [2.43.6](https://github.com/n8n-io/n8n/compare/n8n@2.43.5...n8n@2.43.6) (2026-10-15)
### Bug Fixes
* **core:** Create an XML parser for each webhook request body ([#41113](https://github.com/n8n-io/n8n/issues/41113)) ([ea3ee53](https://github.com/n8n-io/n8n/commit/ea3ee53361503ac55478ac146d8474765ea08574))
<!-- This is an auto-generated description by cubic. -->
<a href="https://cubic.dev/pr/n8n-io/n8n/pull/41153?utm_source=github" target="_blank" rel="noopener noreferrer" data-no-image-dialog="true"><picture><source media="(prefers-color-scheme: dark)" srcset="https://www.cubic.dev/buttons/review-in-cubic-dark.svg"><source media="(prefers-color-scheme: light)" srcset="https://www.cubic.dev/buttons/review-in-cubic-light.svg"><img alt="View guided diff" src="https://www.cubic.dev/buttons/review-in-cubic-light.svg"></picture></a> <a href="https://www.cubic.dev/action/auto-fix/pr/n8n-io/n8n/41153?returnTo=https%3A%2F%2Fgithub.com%2Fn8n-io%2Fn8n%2Fpull%2F41153&source=description" ta
## [2.44.2](https://github.com/n8n-io/n8n/compare/n8n@2.44.1...n8n@2.44.2) (2026-10-15)
### Bug Fixes
* **core:** Create an XML parser for each webhook request body ([#41112](https://github.com/n8n-io/n8n/issues/41112)) ([c52f181](https://github.com/n8n-io/n8n/commit/c52f1817b38cb03758fee45b1109caae12411ffa))
<!-- This is an auto-generated description by cubic. -->
<a href="https://cubic.dev/pr/n8n-io/n8n/pull/41154?utm_source=github" target="_blank" rel="noopener noreferrer" data-no-image-dialog="true"><picture><source media="(prefers-color-scheme: dark)" srcset="https://www.cubic.dev/buttons/review-in-cubic-dark.svg"><source media="(prefers-color-scheme: light)" srcset="https://www.cubic.dev/buttons/review-in-cubic-light.svg"><img alt="View guided diff" src="https://www.cubic.dev/buttons/review-in-cubic-light.svg"></picture></a> <a href="https://www.cubic.dev/action/auto-fix/pr/n8n-io/n8n/41154?returnTo=https%3A%2F%2Fgithub.com%2Fn8n-io%2Fn8n%2Fpull%2F41154&source=description" ta
## Features
- feat(project-files): add core storage and realtime operations (#9016)
- feat(library): 7 Best AI Agents for Recruiting Teams in 2026: Screening, Scheduling, ATS, and Outreach (#9031)
- feat(library): 7 Best Voice AI Agent Builders for Phone Support in 2026 (#9030)
- feat(library): Best AI Agents for IT Helpdesk Teams: 6 Platforms Compared (#9029)
- feat(library): Best AI Agent Platforms for Agencies and Consultants Managing Multiple Clients (#9028)
- feat: establish shared file ownership foundation (#8609)
- feat(library): How to Give AI Agents Live Web Search with the You.com API in Sim (#9007)
- feat(changelog): cover October 13 product releases (#9004)
- feat(mothership): add super-user Plan and knowledge graph settings with benchmark tooling (#8637)
## Improvements
- improvement(docs): standardize integration copy and generated references (#9039)
- improvement(naming): rename sandbox-tasks to sandboxes and standardize file names (#9044)
- improvement(ci): track fla
## NVIDIA Nemotron 3.5 Super
[NVIDIA Nemotron 3.5 Super](https://ollama.com/library/nemotron-3.5-super) is a vision-language model for coding, reasoning, tool use, and document understanding. It pairs a 120B-parameter hybrid Mamba-2/Transformer mixture-of-experts backbone (12B active per token) with a C-RADIO vision encoder, and supports tool calling, an on/off thinking mode, and context up to 1M tokens.
```
ollama run nemotron-3.5-super
```
A compact sibling, [Nemotron 3.5 Bolt](https://ollama.com/library/nemotron-3.5-bolt), distills Super down to 74B parameters (9B active) tuned to fit a 24 GB NVIDIA GPU for everyday agentic tasks, with speculative decoding via its included drafter.
## What's Changed
* Decision models are now available on cloud — use `clef:cloud` and `clef-flash:cloud` to run decision models through ollama.com cloud without pulling them locally
* Continuing an OpenAI-compatible Responses conversation that includes custom tool calls no longer fails wit
## What's Changed
* feat(llm-connections): label TypeSafe adapter as "TypeSafe (System One)" in picker by @hassiebp in https://github.com/langfuse/langfuse/pull/18670
* perf(worker): poll queue metrics once per interval across workers by @Steffen911 in https://github.com/langfuse/langfuse/pull/18672
* ci(notifications): stop sending redundant title, branch, and event to Slack by @nimarb in https://github.com/langfuse/langfuse/pull/18652
* chore(pricing): correct Claude Sonnet 5.5 cache-read price by @langfuse-bot in https://github.com/langfuse/langfuse/pull/18667
* fix(web): Improve tool response handling in sessions by @bezbac in https://github.com/langfuse/langfuse/pull/18640
* fix(web): combine account and organization settings into one page by @nmtrang29 in https://github.com/langfuse/langfuse/pull/18530
* chore(observability): drop unused ClickHouse cost and export heartbeat row metrics by @niklassemmler in https://github.com/langfuse/langfuse/pull/18678
* fix(trace-batchi
## [2.43.5](https://github.com/n8n-io/n8n/compare/n8n@2.43.4...n8n@2.43.5) (2026-10-14)
### Bug Fixes
* **core:** Validate the size argument of the chunk() expression extension ([#40911](https://github.com/n8n-io/n8n/issues/40911)) ([4bf0ece](https://github.com/n8n-io/n8n/commit/4bf0ece7ae665034b3bd6409d0ec0beef4384a57))
### Features
* Show Gemini models by name and list prioritized models first in model pickers ([#41016](https://github.com/n8n-io/n8n/issues/41016)) ([66cece2](https://github.com/n8n-io/n8n/commit/66cece279da3feaab86d773fa0fcefbd9e8b141d))
<!-- This is an auto-generated description by cubic. -->
<a href="https://cubic.dev/pr/n8n-io/n8n/pull/41046?utm_source=github" target="_blank" rel="noopener noreferrer" data-no-image-dialog="true"><picture><source media="(prefers-color-scheme: dark)" srcset="https://www.cubic.dev/buttons/review-in-cubic-dark.svg"><source media="(prefers-color-scheme: light)" srcset="https://www.cubic.dev/buttons/review-in-cubic-light.svg"><img
## [2.44.1](https://github.com/n8n-io/n8n/compare/n8n@2.44.0...n8n@2.44.1) (2026-10-14)
### Features
* Show Gemini models by name and list prioritized models first in model pickers ([#41015](https://github.com/n8n-io/n8n/issues/41015)) ([321b020](https://github.com/n8n-io/n8n/commit/321b0203b73e3351c766cbece6a8f79c800d453f))
<!-- This is an auto-generated description by cubic. -->
<a href="https://cubic.dev/pr/n8n-io/n8n/pull/41040?utm_source=github" target="_blank" rel="noopener noreferrer" data-no-image-dialog="true"><picture><source media="(prefers-color-scheme: dark)" srcset="https://www.cubic.dev/buttons/review-in-cubic-dark.svg"><source media="(prefers-color-scheme: light)" srcset="https://www.cubic.dev/buttons/review-in-cubic-light.svg"><img alt="View guided diff" src="https://www.cubic.dev/buttons/review-in-cubic-light.svg"></picture></a> <a href="https://www.cubic.dev/action/auto-fix/pr/n8n-io/n8n/41040?returnTo=https%3A%2F%2Fgithub.com%2Fn8n-io%2Fn8n%2Fpull%2F41040&source
## What's Changed
### Features
- Add crewai eval command to report where a stopped model broke.
- Use LiteLLM catalog for context windows.
### Bug Fixes
- Fix flow to learn from resumed human feedback and preserve reviewed output types.
- Warn when non-empty apps resolve to zero platform tools.
- Fix flow add-crew to correctly find the package from pyproject.
- Show reasons for deployment failures and fix deploy remove and lockfile checks.
- Fix trigger list and payload requests timeout to 60 seconds.
- Read AMP's source_type during deploy push to ensure ZIP deployments use the new code.
- Decode local JSON files with UTF-8 BOM in rag.
- Allow image input_files for OpenAI GPT-6 Sol/Luna/Astra.
- Emit call completed event for streamed anthropic tool calls.
- Stop JSON-schema models from warning about BaseModel-shadowing fields.
- Preserve native tool aborts and post-hook cache identity.
- Refresh LiteLLM catalog after the version-bump PR merges.
- Repair rotted evaluator regression te
## Features
- feat(knowledge): index Confluence image attachments (#8949)
- feat(projects): enforce Project membership and retire the connector (#8590)
- feat(library): 6 Best AI Agents for Invoice Processing and Finance Reconciliation in 2026 (#8932)
- feat(library): Best AI Agents for Web Scraping and Market Research in 2026 (#8931)
## Improvements
- improvement(search): Confluence and Jira partial-match recall, passage previews, batched MCP reads (#8968)
- improvement(db): cut hot-path connector count and rate-limit lock costs (#8944)
- improvement(perf): bound trash, file-path and attachment reads (#8945)
- improvement(db): disable embedding FTS fastupdate and index delete-action foreign keys (#8943)
- improvement(changelog): add meaningful release coverage (#8942)
## Bug Fixes
- fix(knowledge): key vision checkpoints by title and keep excluded images off while disabled (#8965)
- fix(lifecycle): retain shared resources when creators leave (#8762)
- fix(recently-deleted): dim t
# [2.44.0](https://github.com/n8n-io/n8n/compare/n8n@2.43.0...n8n@2.44.0) (2026-10-13)
### Bug Fixes
* Align execution retry menu and clarify actions ([#40710](https://github.com/n8n-io/n8n/issues/40710)) ([84d2fd6](https://github.com/n8n-io/n8n/commit/84d2fd6f3ae68e973aa49f7e2b1dd5dbf06ad18c))
* Allow per-process crash journal paths ([#37670](https://github.com/n8n-io/n8n/issues/37670)) ([447e939](https://github.com/n8n-io/n8n/commit/447e939b732b778662572098d81f014527e3ffe2))
* **core:** Allow reverting a workflow version that uses credentials the user cannot use ([#40701](https://github.com/n8n-io/n8n/issues/40701)) ([033a9b5](https://github.com/n8n-io/n8n/commit/033a9b512c2d52d05b43d4dd2d0e10634dc36e2a))
* **core:** Apply node type policies to agent node tools ([#39711](https://github.com/n8n-io/n8n/issues/39711)) ([e7d8533](https://github.com/n8n-io/n8n/commit/e7d8533880df0d36438e39a07f0594fcaaaaab55))
* **core:** Apply the credential type policy to credential tests and probes (
## What's Changed
### Features
* feat(web): require Turnstile on password-reset emails by @maxdeichmann in https://github.com/langfuse/langfuse/pull/18584
* feat(ci): bump dependabot to opus5.5 by @nimarb in https://github.com/langfuse/langfuse/pull/18553
* feat(otel): extract media before TypeScript parsing by @sumerman in https://github.com/langfuse/langfuse/pull/18239
* feat(media): add upload pool with concurrency and byte limits by @sumerman in https://github.com/langfuse/langfuse/pull/18629
* feat(worker): per-project trace idle time in tracing_readiness_settings by @LysanderKie in https://github.com/langfuse/langfuse/pull/18595
* feat(topics): configure Topics on the project's own LLM connections by @LysanderKie in https://github.com/langfuse/langfuse/pull/18414
### Fixes
* fix(native): avoid repeated UTF-8 validation during media extraction by @sumerman in https://github.com/langfuse/langfuse/pull/18654
* fix(topics): clean up trace list labels by @nimarb in https
## What's Changed
* feat(web): hint self-hosted v3 deployments to upgrade to Langfuse v4 by @maxdeichmann in https://github.com/langfuse/langfuse/pull/18644
* fix(web): tidy the v3 version menu upgrade hint by @maxdeichmann in https://github.com/langfuse/langfuse/pull/18661
**Full Changelog**: https://github.com/langfuse/langfuse/compare/v3.225.11...v3.225.12
## Features
- feat(changelog): publish the Nebius model update (#8924)
- feat(library): add Stripe billing and failed-payment agent guide
- feat(library): add Stripe payment automation and alert guide
- feat(library): How Do You Build a HubSpot AI Agent for Lead Enrichment and Follow-Up? (#8907)
- feat(library): How Do You Build an AI Agent That Answers Questions From Confluence? (#8919)
- feat(library): AI Agent Version Control and Rollback: Platforms, Process, and Enterprise Checklist (#8918)
- feat(library): How Do You Control AI Agent Spend Across Teams? (#8917)
## Improvements
- perf(tables): read a sorted view's order once per drain and project selected columns in SQL (#8915)
- perf(mothership): stop re-reading and over-serving chat transcripts (#8914)
- perf(logs): order public log pages on an indexed column and retire a superseded document index (#8913)
## Bug Fixes
- fix(chat): name stackless error subclasses and compact stored unfinished tool params (#8928)
- fix(google-
## 1.27.1 (2026-10-12)
<!-- Release notes generated using configuration in .github/release.yml at main -->
## What's Changed
### Bug Fixes
* fix(docker): copy MARS web bridge modules into the image by @VascoSch92 in https://github.com/OpenHands/OpenHands/pull/18343
* fix(themes): make warning feedback theme-aware by @enyst in https://github.com/OpenHands/OpenHands/pull/18333
### Documentation
* docs(verify-openhands): map maintenance 2026-10-12 by @smolpaws in https://github.com/OpenHands/OpenHands/pull/18325
**Full Changelog**: https://github.com/OpenHands/OpenHands/compare/v1.27.0...v1.27.1
---
This PR was generated with [Release Please](https://github.com/googleapis/release-please). See [documentation](https://github.com/googleapis/release-please#release-please).
<!-- AGENT_CANVAS_DOCKER_START -->