Backfill issue -284 — The AI Toolchain

EDITIONS tail grep head diff uniq

The daily firehose — everything the toolchain shipped today, already filtered.

// HOW THIS ISSUE IS MADE

We read every release from the 174 tools on our watchlist at the source — GitHub and GitLab release notes, vendor release pages and changelogs, project blogs and feeds, vendor press releases, and the source code behind the tag. Bug-fix-only releases and non-product newsroom noise are dropped; what's left is summarized down to the new capability, how to try it, and any screenshots or videos the release itself published. Every entry links to the sources it was built from.

VIEW

ISSUE VIEW full issue

Do you prefer this view?

$ tct list   # 11 tools matched

AI & LLM Tooling

◆ AI Coding Agents

Cline

v3.36.0 NOTES

Autonomous coding agent as an SDK, IDE extension, or CLI assistant.

Cline v3.36.0 adds AIhubmix provider, promotes Hooks to GA, enables native tool calls by default, and ships initial CLI samples.

└──▷ GET THIS VERSION

$ git clone --branch v3.36.0 https://github.com/cline/cline.git
# already have the repo? check out this version:
$ git checkout v3.36.0

›Adds AIhubmix as a new AI provider option.
›Promotes Hooks out of feature-flag preview — hooks setting toggle remains available.
›Enables native tool call mode by default for all supported providers.
›Adds native tool call usage tracking via telemetry.
›Adds GCP Vertex as a provider option in remote config schema.

+2 moreshow less

›Filters MCP Marketplace results based on remote configuration, with an info banner shown during marketplace lockdown state.
›Ships initial CLI sample scripts to help practitioners get started with Cline's CLI interface.

└──▷ BREAKING ON UPGRADE

!Native tool calls are now enabled by default; setups that relied on the previous default-off behavior may see changed request formats to compatible providers.

Was this useful?

Charm Crush

v0.15.2 NOTES

Glamourous agentic coding for all

Crush v0.15.2 adds Ctrl+L/Ctrl+M hotkeys to open the model switcher from anywhere.

└──▷ GET THIS VERSION

$ git clone --branch v0.15.2 https://github.com/charmbracelet/crush.git
# already have the repo? check out this version:
$ git checkout v0.15.2

›Adds Ctrl+L and Ctrl+M keyboard shortcuts to open the model switcher without leaving the keyboard.

Was this useful?

Block Goose

v1.13.0 NOTES

an open source, extensible AI agent that goes beyond code suggestions - install, execute, edit, and test with any LLM

Goose v1.13.0 adds Mistral AI provider, newer GitHub Copilot models, platform tools in CLI, and env-var auth config.

└──▷ GET THIS VERSION

$ git clone --branch v1.13.0 https://github.com/block/goose.git
# already have the repo? check out this version:
$ git checkout v1.13.0

›Adds Mistral AI as a supported LLM provider.
›Supports newer GitHub Copilot models.
›Exposes platform tools in the CLI (previously desktop-only).
›Enables environment variables in auth configuration.
›Adds Ctrl-C cancellation of in-flight AI requests from the CLI.

+2 moreshow less

›Adds right-click dock icon for opening a new window on macOS.
›Shows a loading indicator for extensions while they initialize.

└──▷ BREAKING ON UPGRADE

!The JetBrains integration has been removed; any existing JetBrains-based workflows will no longer function after upgrading.

Was this useful?

SST OpenCode

v1.0.37 2 RELEASES · 2025-11-06 NOTES STABLE

The open source coding agent.

OpenCode v1.0.37 adds system theme support and a desktop review flow.

└──▷ GET THIS VERSION

$ git clone --branch v1.0.37 https://github.com/sst/opencode.git
# already have the repo? check out this version:
$ git checkout v1.0.37

›Adds system theme support, allowing the UI to follow the OS light/dark theme preference.
›Adds a desktop review flow for reviewing changes directly within the desktop app.

1 more release in this issue · 2025-11-06

v1.0.36 NOTES STABLE

OpenCode v1.0.36 adds share link support in import and a current-session indicator in the session modal.

└──▷ GET THIS VERSION

$ git clone --branch v1.0.36 https://github.com/sst/opencode.git
# already have the repo? check out this version:
$ git checkout v1.0.36

└──▷ TRY IT

Import a shared session directly from a share link instead of a local file.

$ opencode import <share-link>

›Adds share link support to the import command, enabling direct import of shared sessions via URL.
›Adds a big dot (●) indicator in the session modal to highlight the current active session.

Was this useful?

Zed

v0.211.5 NOTES

Zed v0.211.5 adds live SVG preview on every keypress and relative-path terminal login support for ACP extensions.

└──▷ GET THIS VERSION

$ git clone --branch v0.211.5 https://github.com/zed-industries/zed.git
# already have the repo? check out this version:
$ git checkout v0.211.5

›SVG preview now rerenders on every keypress instead of only on file saves, enabling live feedback while editing SVG files.
›ACP extensions can now run terminal login commands using relative paths.

Was this useful?

◆ AI Agent Frameworks

LangChain

langchain==1.0.4 NOTES

LangChain 1.0.4 adds model-profiles as an optional dependency.

└──▷ GET THIS VERSION

$ git clone --branch langchain==1.0.4 https://github.com/langchain-ai/langchain.git
# already have the repo? check out this version:
$ git checkout langchain==1.0.4

›Adds model-profiles as an optional dependency for LangChain, enabling model profile support.

Was this useful?

PydanticAI

v1.11.1 NOTES

PydanticAI v1.11.1 adds FallbackModel support for Native output mode and ModelProfile.default_structured_output_mode

└──▷ GET THIS VERSION

$ git clone --branch v1.11.1 https://github.com/pydantic/pydantic-ai.git
# already have the repo? check out this version:
$ git checkout v1.11.1

›Adds ModelProfile.default_structured_output_mode support to FallbackModel, enabling native output mode control when falling back across models.

Was this useful?

◆ Local LLM Runtimes

SGLang

v0.5.5 NOTES

SGLang v0.5.5 adds Kimi-K2, MiniMax-M2, NVILA model support, mTLS for router workers, MCP connection pooling, and ~10% B200 MoE perf gains.

└──▷ GET THIS VERSION

$ git clone --branch v0.5.5 https://github.com/sgl-project/sglang.git
# already have the repo? check out this version:
$ git checkout v0.5.5

└──▷ TRY IT

Run MiniMax-M2 with the new model support on a multi-GPU node.

$ python -m sglang.launch_server --model-path MiniMaxAI/MiniMax-M2 --tp 8

›Adds mTLS support for router-to-worker communication, securing inter-component traffic in multi-node deployments.
›Adds MCP Manager to the router with connection pooling, tool inventory, and proxy support.
›Adds flashinfer_trtllm MoE runner backend, delivering ~10% throughput improvement on B200 FP8 DeepSeek workloads.
›Adds deterministic inference support for DeepSeek architecture models on a single GPU.
›Adds DeepGEMM support for deterministic inference.

+10 moreshow less

›Adds DeepSeek-R1 w4a8 low-latency DeepEP support.
›Adds CUDA graph support for trtllm_mla attention backend spec decoding, improving speculative decoding throughput.
›Adds support for MiniMax-M2 model.
›Adds support for NVILA and NVILA Lite vision-language models.
›Adds support for Kimi-K2-Thinking model.
›Adds Llama 4 + TRTLLM MHA support.
›Adds ResponsesContext to the gRPC router responses API with improved error propagation.
›Adds MHA chunked-prefix optimization that merges prefix and extend KV cache to run MHA in a single pass.
›Removes the deprecated --enable-beta-spec CLI argument.
›Decouples GGUF quantization implementation from the vLLM dependency, and adds gguf as a first-class project dependency for CPU/XPU targets.

└──▷ BREAKING ON UPGRADE

!The --enable-beta-spec argument has been removed; invocations passing this flag will fail on upgrade.

Was this useful?

oobabooga's Text Generation WebUI (textgen)

v3.17 NOTES

v3.17 adds Qwen3-VL support via llama.cpp update and bumps exllamav3 to 0.0.12.

└──▷ GET THIS VERSION

$ git clone --branch v3.17 https://github.com/oobabooga/textgen.git
# already have the repo? check out this version:
$ git checkout v3.17

›Supports Qwen3-VL multimodal models through the updated llama.cpp backend.
›Updates exllamav3 backend to version 0.0.12.
›Adds weights_only=True to torch.load in Training_PRO to harden model checkpoint loading.

Was this useful?

Other / Uncategorized

◆ AI OBSERVABILITY

Langfuse

v3.127.0 NOTES

Langfuse v3.127.0 adds collapsible system prompts, hover name display in trace trees, and Python dict parsing in the trace UI.

└──▷ GET THIS VERSION

$ git clone --branch v3.127.0 https://github.com/langfuse/langfuse.git
# already have the repo? check out this version:
$ git checkout v3.127.0

›Collapses long system prompts in the trace UI to reduce visual noise when inspecting verbose prompt templates.
›Displays the span or step name on hover in the trace tree for faster identification without expanding nodes.
›Parses Python dict syntax in the trace UI, enabling proper rendering of Python-style objects in trace inputs and outputs.
›Adds dual writes for bookmarked events in the events table to keep bookmark state consistent across the new table schema.

Was this useful?

Weights & Biases Weave

v0.52.15 NOTES

Weave v0.52.15 adds batch object creation and OTEL span op generation.

└──▷ GET THIS VERSION

$ git clone --branch v0.52.15 https://github.com/wandb/weave.git
# already have the repo? check out this version:
$ git checkout v0.52.15

›Implements obj_create_batch for creating multiple objects in a single batch operation.
›Adds op creation for OTEL spans, enabling tracing operations to be generated from OpenTelemetry span data.

Was this useful?

// END OF TRANSMISSION

The AI Toolchain · curated daily · November 6, 2025
Releases & features from the open-source and commercial AI tools that practitioners actually run.

The AI Toolchain — issue -284, November 6, 2025