Heads up This site is currently under heavy development.
Subscribe Get it delivered — the daily firehose, filtered to the tools you run, plus the documentation changes vendors never announce. Compare plans →

The AI Toolchain — issue -249, December 11, 2025

THE AI TOOLCHAIN NO. -249
Tail
THE DAILY RELEASE FIREHOSE
PUBLISHED DECEMBER 11, 2025 · EVERY WEEKDAY
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   # 16 tools matched
AI & LLM Tooling
◆  AI Agent Frameworks

Agno (formerly Phidata)

Sources Release notes → v2.3.11 NOTES

Agno v2.3.11 adds OpenAI-specific fields to RunOutput and RunCompletedEvent for streaming and non-streaming access.

└──▷ GET THIS VERSION
$ git clone --branch v2.3.11 https://github.com/agno-agi/agno.git
# already have the repo? check out this version:
$ git checkout v2.3.11
  • Adds new OpenAI-specific fields to RunOutput and RunCompletedEvent classes via response_provider_data, exposing provider data in both streaming and non-streaming cases.
Was this useful?

AutoGPT

Sources Release notes → autogpt-platform-beta-v0.6.40 NOTES

AutoGPT Platform v0.6.40 adds GitHub webhook triggers, new spreadsheet blocks, execution accuracy alerts, and unified Google Sheets/Drive credentials.

└──▷ GET THIS VERSION
$ git clone --branch autogpt-platform-beta-v0.6.40 https://github.com/Significant-Gravitas/AutoGPT.git
# already have the repo? check out this version:
$ git checkout autogpt-platform-beta-v0.6.40
  • Adds 4 new GitHub webhook trigger blocks for event-driven agent automation.
  • Adds many new spreadsheet blocks expanding data manipulation capabilities in the flow builder.
  • Adds an execution accuracy alert system to surface agent output quality issues.
  • Improves Google Sheets and Google Drive integration with unified credentials, replacing separate credential flows.
  • Adds a file input widget with variants and base64 support to flow editor nodes.
+10 moreshow less
  • Adds host-scoped credentials support to CredentialField for per-host credential management.
  • Adds builder search history so previously searched terms are recalled in the flow builder.
  • Adds an 'Agent Output Demo' field to the marketplace submission form for showcasing agent behavior.
  • Adds templates and triggers to the new Library page view.
  • Implements a new actions sidebar with summary panel in the UI.
  • Switches copied graph storage from local storage to clipboard for cross-session paste support.
  • Adds an expandable text input modal for editing long field values in the flow editor.
  • Displays graph validation errors inline on node fields instead of as separate notifications.
  • Adds automatic collision resolution for flow editor nodes to prevent overlapping on the canvas.
  • Completes backend support for onboarding task tracking.
Was this useful?

LangChain

Sources Release notes → langchain-openai==1.1.2 NOTES

langchain-openai 1.1.2 adds a strict flag to ProviderStrategy structured output.

└──▷ GET THIS VERSION
$ git clone --branch langchain-openai==1.1.2 https://github.com/langchain-ai/langchain.git
# already have the repo? check out this version:
$ git checkout langchain-openai==1.1.2
  • Adds strict flag to ProviderStrategy structured output, enabling strict schema enforcement when using provider-based structured output.
Was this useful?

OpenAI Agents SDK

Sources Release notes → v0.6.3 NOTES

OpenAI Agents SDK v0.6.3 preserves logprobs from the chat completions API in ModelResponse.

└──▷ GET THIS VERSION
$ git clone --branch v0.6.3 https://github.com/openai/openai-agents-python.git
# already have the repo? check out this version:
$ git checkout v0.6.3
  • Preserves logprobs data from the chat completions API in ModelResponse, making token-level probability information available to callers downstream.
Was this useful?

PydanticAI

Sources Release notes → v1.30.0 NOTES

PydanticAI v1.30.0 adds CerebrasModel, prompt caching options for OpenAI, and multi-modal output in LLMJudge.

└──▷ GET THIS VERSION
$ git clone --branch v1.30.0 https://github.com/pydantic/pydantic-ai.git
# already have the repo? check out this version:
$ git checkout v1.30.0
  • Adds CerebrasModel as a new supported LLM provider integration.
  • Adds prompt caching options to OpenAIChatModelSettings for controlling OpenAI prompt cache behavior.
  • Supports multi-modal output in LLMJudge evaluations.
Was this useful?
◆  AI Coding Agents

Cline

Sources Release notes → v3.41.0 NOTES

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

Cline v3.41.0 adds web search/fetch tools, OpenAI Responses API, parallel tool calling, hooks in CLI, and new model support including GPT-5.2 and Amazon Nova 2 Lite.

└──▷ GET THIS VERSION
$ git clone --branch v3.41.0 https://github.com/cline/cline.git
# already have the repo? check out this version:
$ git checkout v3.41.0
  • Adds web search capability with a new websearch tool and minimal search usage UI, enabling Cline to query the web during tasks.
  • Adds updated webfetch tool with a new handler for fetching web content, accessible via the new Web Tools features menu.
  • Adds OpenAI Responses API support with Codex model compatibility.
  • Adds experimental parallel tool calling support, allowing multiple tools to be invoked simultaneously.
  • Enables hooks in the CLI, bringing hook lifecycle events to headless/CLI workflows.
+15 moreshow less
  • Adds new PreCompact hook, triggered before context compaction occurs.
  • Adds LiteLLM provider to remote config, with API key configurable via remote config.
  • Adds Vertex provider to remote config.
  • Adds native tool calling support for DeepSeek V3.2.
  • Adds DeepSeek V3.2 to the static models list.
  • Adds Amazon Nova 2 Lite model support.
  • Adds GPT-5.2 model support in the model picker.
  • Adds new Mistral model definitions with proxy support.
  • Redesigns model settings page with compact info layout and an Advanced section.
  • Improves chat modal model picker UI.
  • Adds dismiss functionality to the banner carousel.
  • Shows MCP messages in CLI output.
  • Moves standalone terminal implementation to TypeScript.
  • Adds welcome screen UI enhancements.
  • Sets restrictive file permissions on secrets.json for improved secret storage security.
Was this useful?

Continue

Sources Release notes → v1.2.12-vscode NOTES

Continue v1.2.12 adds headless CLI mode, grep/find/findstr search fallback, Gemini thought signatures, org profile selection, and auto-approval of parallel read-only tools.

└──▷ GET THIS VERSION
$ git clone --branch v1.2.12-vscode https://github.com/continuedev/continue.git
# already have the repo? check out this version:
$ git checkout v1.2.12-vscode
└──▷ TRY IT
Run the Continue CLI agent in a headless CI environment where no TTY is available.
$ continue --headless
  • Removes TTY requirement for the CLI in headless mode, enabling non-interactive/CI environments to run the CLI without a terminal.
  • Adds grep, find, and findstr as alternative search strategies in the CLI when ripgrep is unavailable.
  • Adds support for Gemini thought signatures in streaming responses via the google/genai SDK, now used for all Gemini and Vertex streaming.
  • Adds org profile selection on sign-in, letting users choose their organization during authentication.
  • Auto-approves parallel read-only built-in tools, reducing interruptions during agent runs.
+11 moreshow less
  • Adds Supabase-specific OAuth scopes to the mcp-remote fallback for MCP connections.
  • Adds agent error reporting to the API in serve mode.
  • Adds agent session metadata tracking and reporting to the control plane.
  • Increases terminal command timeout from 30 seconds to 120 seconds, unblocking longer-running agent tasks.
  • Adds auto-continuation after context compaction in chat streaming, so long sessions resume without manual intervention.
  • Adds support for renaming a rule before creation in the rules UI.
  • Adds support for claude-opus-4 and claude-sonnet-4.5 model updates.
  • Adds a JCEF reload action for the JetBrains extension.
  • Adds Klavis AI Strata to the MCP Cookbook.
  • Adds support for grok-4.1-fast and gpt-5.1 models.
  • Adds support for gemini-3-pro-preview.
Was this useful?

Charm Crush

Sources Release notes → v0.23.0 NOTES

Glamourous agentic coding for all

Crush v0.23.0 adds GPT-5.2 and GPT-5.1 Codex Max support, plus MCP disabled_tools config.

└──▷ GET THIS VERSION
$ git clone --branch v0.23.0 https://github.com/charmbracelet/crush.git
# already have the repo? check out this version:
$ git checkout v0.23.0
└──▷ USE IT
Suppress noisy or dangerous MCP tools from being offered to the model without removing the whole server.
yaml
# In your MCP server config block, add:
disabled_tools:
  - bash
  - write_file
  • Adds support for OpenAI GPT-5.2 as a selectable model.
  • Adds support for GPT-5.1 Codex Max model.
  • New disabled_tools list in MCP config lets you suppress specific MCP tools without removing the server.
  • Shows a / hint in the TUI editor when the input field is empty, surfacing command discovery.
Was this useful?

OpenAI Codex CLI

Sources Release notes → rust-v0.71.0 NOTES

Lightweight coding agent that runs in your terminal

Codex CLI rust-v0.71.0 adds gpt-5.2 support, shell snapshots, long-snapshot warnings, and an elevated Sandbox 2.

└──▷ GET THIS VERSION
$ git clone --branch rust-v0.71.0 https://github.com/openai/codex.git
# already have the repo? check out this version:
$ git checkout rust-v0.71.0
  • Adds gpt-5.2 as a selectable model, OpenAI's latest frontier model with improvements across knowledge, reasoning, and coding.
  • Shows the default model in the model picker so practitioners can see at a glance what will be used.
  • Adds shell snapshots for shell commands, capturing environment state for reproducibility.
  • Adds warnings when snapshots grow too long, surfacing potential performance or context issues early.
  • Introduces Elevated Sandbox 2, an upgraded sandboxing environment.
Was this useful?

SST OpenCode

Sources Release notes → v1.0.150 3 RELEASES · 2025-12-11 NOTES STABLE

The open source coding agent.

OpenCode v1.0.150 adds super modifier (Command/Windows key) support in keybinds.

└──▷ GET THIS VERSION
$ git clone --branch v1.0.150 https://github.com/sst/opencode.git
# already have the repo? check out this version:
$ git checkout v1.0.150
  • Supports the super modifier (Command/Windows key) in keybinds, enabling shortcuts like super+z for undo and super+shift+z for redo.
2 more releases in this issue · 2025-12-11
v1.0.147 NOTES STABLE

OpenCode v1.0.147 adds custom keybindings for all textarea actions, word-navigation shortcuts, and ripgrep in Docker.

└──▷ GET THIS VERSION
$ git clone --branch v1.0.147 https://github.com/sst/opencode.git
# already have the repo? check out this version:
$ git checkout v1.0.147
  • Adds support for custom keybindings for all textarea actions in the input area via config.
  • Adds Ctrl+Right/Left for word navigation, Ctrl+D for character deletion, and Alt+D for word deletion as default input shortcuts.
  • Includes ripgrep in the official Docker image, enabling fast code search without extra installation.
v1.0.144 NOTES STABLE

OpenCode v1.0.144 adds Cerebras integration, models.dev support, and enables icon discovery by default in the Tauri app.

└──▷ GET THIS VERSION
$ git clone --branch v1.0.144 https://github.com/sst/opencode.git
# already have the repo? check out this version:
$ git checkout v1.0.144
  • Adds Cerebras integration header identifying requests with the opencode identifier.
  • Adds support for the models.dev model family.
  • Enables icon discovery feature by default in the Tauri app.
  • TUI now displays only the count of active MCP servers.
  • Tauri app now automatically selects an available port instead of defaulting to 4096.
+4 moreshow less
  • Uses |- formatting for intermediate sub-agent steps, improving multi-agent output readability.
  • Improves interleaved thinking functionality.
  • Sets Gemini thinking level to high by default.
  • Improves compaction prompt for context management.
└──▷ BREAKING ON UPGRADE
  • !Gemini thinking level is now set to high by default, which may change model behavior and increase token usage for existing Gemini workflows.
Was this useful?

Alibaba Qwen Code

Sources Release notes → v0.4.1-nightly.20251211.a02c4b27 NOTES

Qwen Code v0.4.1-nightly adds a VSCode extension and a cleaner input prompt UI for easier copy/paste.

└──▷ GET THIS VERSION
$ git clone --branch v0.4.1-nightly.20251211.a02c4b27 https://github.com/QwenLM/qwen-code.git
# already have the repo? check out this version:
$ git checkout v0.4.1-nightly.20251211.a02c4b27
  • Adds a VSCode Extension, bringing Qwen Code capabilities directly into the editor.
  • Removes vertical borders from the input prompt UI, making it easier to copy and paste prompts.
Was this useful?

Zed

Sources Release notes → v0.216.1 NOTES

Zed v0.216.1 adds GPT-5.2 support and shows extension-defined agent display names in the menu.

└──▷ GET THIS VERSION
$ git clone --branch v0.216.1 https://github.com/zed-industries/zed.git
# already have the repo? check out this version:
$ git checkout v0.216.1
  • Adds support for OpenAI's GPT-5.2 as an available model.
  • Renders agent display names defined by extensions in the agent selection menu.
Was this useful?
◆  Local LLM Runtimes

llama.cpp

Sources Release notes → b7353 2 RELEASES · 2025-12-11 NOTES STABLE

llama.cpp b7353 enables Jinja templating by default in the CLI.

└──▷ GET THIS VERSION
$ git clone --branch b7353 https://github.com/ggml-org/llama.cpp.git
# already have the repo? check out this version:
$ git checkout b7353
  • Jinja templating is now enabled by default in the CLI (previously required explicit opt-in via common/arg.cpp).
└──▷ BREAKING ON UPGRADE
  • !Linux releases will move from .zip to .tar.gz archives — deployment scripts that unzip Linux builds will need to be updated.
1 more release in this issue · 2025-12-11
b7352 NOTES STABLE

llama-server gains INI-based per-model presets and recursive GGUF directory loading for multi-model deployments.

└──▷ GET THIS VERSION
$ git clone --branch b7352 https://github.com/ggml-org/llama.cpp.git
# already have the repo? check out this version:
$ git checkout b7352
└──▷ USE IT
Define per-model settings in the auto-generated config.ini using LLAMA_ARG_* keys under a section named after each model.
ini
[my-model]
LLAMA_ARG_N_GPU_LAYERS=35
LLAMA_ARG_FLASH_ATTN=1
  • Adds INI-based per-model configuration presets to llama-server via config.ini, with auto-creation and template generation; keys must follow the LLAMA_ARG_* naming convention and section names starting with LLAMA_ARG_* are rejected.
  • Adds LLAMA_ARG_* environment variable passthrough so global settings reach child model processes without CLI duplication — children receive only minimal args (executable, model, port, alias).
  • Adds LLAMA_ARG_HOST and CHILD_ADDR support for controlling child process network binding in multi-model router mode.
  • Replaces flat --models-dir directory scan with recursive traversal using nested vendor/model layouts (e.g. vendor/model/*.gguf); model names now reflect the relative path within --models-dir rather than just the filename.
└──▷ BREAKING ON UPGRADE
  • !Linux releases will move from .zip archives to .tar.gz archives — deployment scripts that fetch or unpack Linux release artifacts will need to be updated.
Was this useful?
Other / Uncategorized
◆  AI OBSERVABILITY

Arize Phoenix

Sources Release notes → arize-phoenix-client-v1.27.0 NOTES

Phoenix client v1.27.0 adds span annotation notes via Python client and REST endpoint, plus LDAP authentication support.

└──▷ GET THIS VERSION
$ git clone --branch arize-phoenix-client-v1.27.0 https://github.com/Arize-ai/phoenix.git
# already have the repo? check out this version:
$ git checkout arize-phoenix-client-v1.27.0
  • Adds a span notes REST endpoint, enabling programmatic creation and retrieval of notes attached to individual spans.
  • Adds a span note function to the Python client library for annotating spans with freeform notes.
  • Adds Lightweight Directory Access Protocol (LDAP) support for authentication.
Was this useful?
◆  VECTOR DB RAG

Milvus

Sources Release notes → v2.5.24 NOTES

Milvus 2.5.24 improves BM25 sparse vector search load performance and cluster load balancing stability.

└──▷ GET THIS VERSION
$ git clone --branch v2.5.24 https://github.com/milvus-io/milvus.git
# already have the repo? check out this version:
$ git checkout v2.5.24
  • Optimizes loading performance of BM25 statistics for sparse vector search, reducing startup latency for relevant workloads.
  • Refines cluster load balancing by excluding stopping nodes from resource groups, improving stability under node churn.
Was this useful?

Weaviate

Sources Release notes → v1.33.10 2 RELEASES · 2025-12-11 NOTES STABLE

Weaviate v1.33.10 adds maintenance mode for gRPC and BaseURL support for Cohere's reranker module.

└──▷ GET THIS VERSION
$ git clone --branch v1.33.10 https://github.com/weaviate/weaviate.git
# already have the repo? check out this version:
$ git checkout v1.33.10
  • Adds BaseURL setting to the Cohere reranker module, enabling use of custom or self-hosted Cohere reranker endpoints.
  • Adds a maintenance mode interceptor to the gRPC server, allowing the server to reject requests during maintenance windows.
1 more release in this issue · 2025-12-11
v1.32.22 NOTES STABLE

Weaviate v1.32.22 adds BaseURL support for Cohere's reranker module and a maintenance mode interceptor for the gRPC server.

└──▷ GET THIS VERSION
$ git clone --branch v1.32.22 https://github.com/weaviate/weaviate.git
# already have the repo? check out this version:
$ git checkout v1.32.22
  • Adds BaseURL setting to the Cohere reranker module, allowing practitioners to point the reranker at a custom or self-hosted Cohere endpoint.
  • Adds a maintenance mode interceptor to the gRPC server, enabling controlled rejection of requests during maintenance windows.
Was this useful?
my-toolchain — 0 tools
paste an install list to detect your tools

A brew list, a Brewfile, requirements.txt, a Dockerfile — or just the product names, free-form. Nothing leaves your browser.

    browse all tools →