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 -340, September 10, 2025

THE AI TOOLCHAIN NO. -340
Tail
THE DAILY RELEASE FIREHOSE
PUBLISHED SEPTEMBER 10, 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   # 14 tools matched
AI & LLM Tooling
◆  AI Agent Frameworks

Agno (formerly Phidata)

Sources Release notes → v2.0.3 NOTES

Agno v2.0.3 adds WorkflowTools, MemoryTools, Gemini TTS, and custom encoding support for reader classes.

└──▷ GET THIS VERSION
$ git clone --branch v2.0.3 https://github.com/agno-agi/agno.git
# already have the repo? check out this version:
$ git checkout v2.0.3
  • Adds WorkflowTools class, usable as a tool for agents and teams to run workflows from within an agent, with reasoning support via think/analyze.
  • Adds MemoryTools class, usable as a tool for agents and teams to add, update, and delete memories, with reasoning support via think/analyze.
  • Adds encoding parameter support on the reader class, allowing custom text encoding to be passed when reading documents.
  • Adds support for additional kwargs on AgentOS API run endpoints, enabling extra arguments when running an agent, team, or workflow via AgentOS.
  • Adds Gemini Text to Speech (TTS) support.
Was this useful?

AutoGPT

Sources Release notes → autogpt-platform-beta-v0.6.26 NOTES

AutoGPT Platform adds Bannerbear, AI Image Customizer, and Gmail Draft Reply blocks alongside drag-and-drop canvas and sub-agent approval.

└──▷ GET THIS VERSION
$ git clone --branch autogpt-platform-beta-v0.6.26 https://github.com/Significant-Gravitas/AutoGPT.git
# already have the repo? check out this version:
$ git checkout autogpt-platform-beta-v0.6.26
  • Adds Bannerbear API block for applying text overlays on images directly in the flow builder.
  • Adds AI Image Customizer block powered by Google's Nano Banana for image manipulation.
  • Adds Gmail Draft Reply block for creating draft email replies from within an agent flow.
  • Implements sub-agent approval support, allowing human-in-the-loop gating of sub-agent execution.
  • Adds API key authentication support in Swagger UI for direct API testing.
+11 moreshow less
  • Adds drag-and-drop functionality from the blocks menu onto the flow canvas.
  • Adds editable node titles with a hover pencil icon in the builder.
  • Adds graph search functionality to the flow builder.
  • Adds recommended run schedule for agent execution.
  • Adds an admin button to the user profile dropdown.
  • Enables 'Run Again' to work with saved credentials.
  • Enables creation of presets directly from past runs.
  • New run agent modal and redesigned block menu improve the builder interface.
  • New agent library run page with timezone display.
  • Enhanced output rendering system for agent runs.
  • Adds comprehensive Block SDK guide to documentation.
Was this useful?

CrewAI

Sources Release notes → 0.186.0 NOTES

Framework for orchestrating role-playing, autonomous AI agents. By fostering collaborative intelligence, CrewAI empowers agents to work together seamlessly, tackling complex tasks.

CrewAI 0.186.0 adds Qdrant RAG support, partial flow resumability, auto-injection of trigger payloads, and a new config reset command.

└──▷ GET THIS VERSION
$ git clone --branch 0.186.0 https://github.com/crewAIInc/crewAI.git
# already have the repo? check out this version:
$ git checkout 0.186.0
└──▷ TRY IT
Reset stored auth/API tokens without being prompted for email — useful when rotating credentials in CI.
$ crewai config reset
  • Adds support for crewai_trigger_payload auto-injection into crew/flow executions.
  • Introduces a centralized RAG configuration system with optional imports, including Qdrant as a new RAG provider.
  • Implements generic clients for ChromaDB and Qdrant storage backends.
  • Enables partial flow resumability, allowing interrupted flows to resume mid-execution.
  • Displays task names in verbose output for clearer runtime observability.
+3 moreshow less
  • Introduces crewai config reset command for resetting authentication tokens.
  • Adds additional parameters to Flow.start() methods for more flexible flow invocation.
  • Introduces centralized configuration for embedding types and a base embedding client.
└──▷ BREAKING ON UPGRADE
  • !A deprecation warning is now emitted for Task.max_retries; callers using this field should migrate before it is removed.
Was this useful?

LangChain

Sources Release notes → langchain-qdrant==0.2.1 3 RELEASES · 2025-09-10 NOTES STABLE

langchain-qdrant 0.2.1 adds similarity_search_with_score_by_vector() to QdrantVectorStore and enables as_retriever in sparse-only mode.

└──▷ GET THIS VERSION
$ git clone --branch langchain-qdrant==0.2.1 https://github.com/langchain-ai/langchain.git
# already have the repo? check out this version:
$ git checkout langchain-qdrant==0.2.1
└──▷ USE IT
Retrieve the top-k most similar documents with scores by passing a raw query vector — useful when you already have embeddings computed upstream.
python
results = qdrant_store.similarity_search_with_score_by_vector(embedding=[0.12, 0.34, ...], k=5)
  • Adds similarity_search_with_score_by_vector() method to QdrantVectorStore, enabling direct vector-based similarity search with relevance scores.
  • Enables as_retriever() to work without embeddings when operating in SPARSE mode, so sparse-only pipelines no longer require a dense embedding model.
2 more releases in this issue · 2025-09-10
langchain-openai==0.3.33 NOTES STABLE

langchain-openai 0.3.33 adds web_search to the supported OpenAI tools list.

└──▷ GET THIS VERSION
$ git clone --branch langchain-openai==0.3.33 https://github.com/langchain-ai/langchain.git
# already have the repo? check out this version:
$ git checkout langchain-openai==0.3.33
  • Adds web_search to the OpenAI tools list, enabling web search as a callable tool in OpenAI-backed chains and agents.
langchain-core==0.3.76 NOTES STABLE

langchain-core 0.3.76 adds id field to Document filters, AWS Bedrock document blocks, multi-format PromptTemplates, and OpenAI web_search tool support.

└──▷ GET THIS VERSION
$ git clone --branch langchain-core==0.3.76 https://github.com/langchain-ai/langchain.git
# already have the repo? check out this version:
$ git checkout langchain-core==0.3.76
  • Adds id field to Document objects passed to the filter callback during InMemoryVectorStore similarity search, enabling filter logic that references document identity.
  • Supports AWS Bedrock document content blocks in msg_content_output, expanding the message content types that can be processed from Bedrock responses.
  • Supports adding PromptTemplates with formats other than f-string, allowing templates using alternative formatting styles to be composed and added together.
  • Adds web_search to the OpenAI tools list, making it available for selection when building OpenAI-backed tool-calling chains.
Was this useful?

PydanticAI

Sources Release notes → v1.0.3 NOTES

PydanticAI v1.0.3 adds sequential tool call context manager, AG-UI callbacks, and Google model seed support

└──▷ GET THIS VERSION
$ git clone --branch v1.0.3 https://github.com/pydantic/pydantic-ai.git
# already have the repo? check out this version:
$ git checkout v1.0.3
└──▷ USE IT
Force an agent to execute tools one at a time — useful when tools have side effects that must not run concurrently.
python
with agent.sequential_tool_calls():
    result = await agent.run('Book a flight then send a confirmation email')
Inspect the full AgentRunResult after an AG-UI run completes, e.g. to log structured output or trigger downstream actions.
python
async def handle_complete(result: AgentRunResult) -> None:
    print(result.output)

await agent.run_as_agui(prompt, on_complete=handle_complete)
Pin a Google model to a fixed random seed so repeated runs return deterministic results during testing.
python
from pydantic_ai.settings import ModelSettings

result = await agent.run('Summarize this doc', model_settings=ModelSettings(seed=42))
  • Adds agent.sequential_tool_calls() context manager to enforce sequential (non-parallel) tool execution within an agent run.
  • Adds on_complete callback to AG-UI functions, providing access to AgentRunResult when a run finishes.
  • Supports ModelSettings.seed in GoogleModel for reproducible outputs.
  • Supports NativeOutput with FunctionModel, enabling native structured output in function-backed models.
  • Sends AG-UI thinking start and end events, surfacing model reasoning steps to AG-UI consumers.
+3 moreshow less
  • Includes thinking parts in subsequent model requests to improve performance and cache hit rates.
  • Raises a clear error when WebSearchTool is used with OpenAIChatModel and an unsupported model, rather than failing silently.
  • Supports models that return output tool args as {"response": "<JSON string>"}, broadening compatibility with non-standard model response formats.
Was this useful?

Microsoft Semantic Kernel

Sources Release notes → dotnet-1.65.0 NOTES

Semantic Kernel dotnet-1.65.0 adds Deep Research Tool for AzureAIAgent, reasoning support for OpenAI Responses Agents, and new AnnotationContent fields.

└──▷ GET THIS VERSION
$ git clone --branch dotnet-1.65.0 https://github.com/microsoft/semantic-kernel.git
# already have the repo? check out this version:
$ git checkout dotnet-1.65.0
  • Adds container_id and filename fields to the AnnotationContent class in the Python library.
  • Adds Deep Research Tool support for AzureAIAgent in the Python connector.
  • Adds reasoning support for OpenAI Responses Agents (GPT-5, o4-mini, o3) in the Python library.
  • Adds framework name into the UserAgent header for the Bedrock integration in Python.
  • Adds thread message ID exposure during streaming for AzureAIAgent in Python.
Was this useful?
◆  AI Coding Agents

Continue

Sources Release notes → v1.2.3-vscode NOTES

Continue v1.2.3 adds Tetrate Agent Router, MCP SSL verification, SSE/streamable-HTTP for MCP, on-prem proxy in CLI, and pause/resume support.

└──▷ GET THIS VERSION
$ git clone --branch v1.2.3-vscode https://github.com/continuedev/continue.git
# already have the repo? check out this version:
$ git checkout v1.2.3-vscode
  • Adds sse and streamable http transport support for MCP in the CLI (feat(cli): support sse and streamable http for mcp).
  • Adds SSL verification for MCP connections.
  • Adds Tetrate Agent Router Service as a new model provider.
  • Supports on-premises proxy in the CLI.
  • Adds pause and resume capability in the CLI (cn).
+3 moreshow less
  • Adds image paste support in the CLI.
  • Warns users when running the CLI from their home directory.
  • Keeps the assistant menu open when clicking reload.
Was this useful?

Charm Crush

Sources Release notes → v0.7.9 NOTES

Glamourous agentic coding for all

Crush v0.7.9 cuts LSP file descriptors by 99% for dramatically faster multi-LSP workflows.

└──▷ GET THIS VERSION
$ git clone --branch v0.7.9 https://github.com/charmbracelet/crush.git
# already have the repo? check out this version:
$ git checkout v0.7.9
  • Optimizes LSP file watcher with 99% reduction in file descriptors, improving performance when using multiple LSPs.
Was this useful?

OpenAI Codex CLI

Sources Release notes → rust-v0.33.0 2 RELEASES · 2025-09-10 NOTES STABLE

Lightweight coding agent that runs in your terminal

Codex CLI 0.33.0 adds a custom Markdown renderer, image support in messages, and keyboard word-deletion.

└──▷ GET THIS VERSION
$ git clone --branch rust-v0.33.0 https://github.com/openai/codex.git
# already have the repo? check out this version:
$ git checkout rust-v0.33.0
  • New custom Markdown renderer replaces tui_markdown for improved in-terminal rendering.
  • Supports images in UserMessageEvent, enabling image attachments in chat messages.
  • Adds alt+delete keybinding to delete the word to the right of the cursor (delete_forward_word).
  • Sets a user-agent suffix when Codex is used as an MCP server.
  • Introduces rollout items for staged feature delivery.
└──▷ BREAKING ON UPGRADE
  • !The POSIX unification and snapshot sessions feature (introduced in #3179) has been reverted; any workflows depending on that behavior will no longer work.
1 more release in this issue · 2025-09-10
rust-v0.32.0 NOTES STABLE

Codex 0.32.0 adds Alt+Delete word deletion, ArchiveConversation API, rollout groundwork, and broader apply_patch OSS model compatibility.

└──▷ GET THIS VERSION
$ git clone --branch rust-v0.32.0 https://github.com/openai/codex.git
# already have the repo? check out this version:
$ git checkout rust-v0.32.0
  • Adds Alt+Delete keybinding in the TUI to delete the next word to the right of the cursor.
  • Adds ArchiveConversation request type to the client protocol for conversation management.
  • Includes rollout_path in NewConversationResponse as initial groundwork for rollout support.
  • Expands apply_patch compatibility to OSS models from gpt-oss providers using alternative naming conventions (e.g. openai/gpt-oss-*).
  • Sets a user-agent suffix when Codex acts as an MCP server.
+1 moreshow less
  • Re-exports InitialHistory from conversation_manager and generates additional TypeScript types, including returning conversation ID with ConversationSummary.
└──▷ BREAKING ON UPGRADE
  • !The config.responses_originator_header_internal_override config key is replaced by the CODEX_INTERNAL_ORIGINATOR_OVERRIDE_ENV_VAR environment variable; any existing config using the old key will no longer work.
Was this useful?

Zed

Sources Release notes → v0.203.4 NOTES

Zed v0.203.4 adds native macOS window tabbing, perceptual font rendering, MCP timeout config, and new editor settings.

└──▷ GET THIS VERSION
$ git clone --branch v0.203.4 https://github.com/zed-industries/zed.git
# already have the repo? check out this version:
$ git checkout v0.203.4
└──▷ USE IT
Enable native macOS window tabbing so Zed merges windows like other macOS apps when full-screened.
json
// In your Zed settings.json:
{ "use_system_window_tabs": true }
Control how many surrounding context lines appear when viewing search results or references in a multibuffer.
json
{ "excerpt_context_lines": 5 }
  • Adds use_system_window_tabs setting to enable native macOS window tabbing, merging windows the same way macOS does (by default when full-screened).
  • Adds excerpt_context_lines setting to configure the number of context lines shown in a multibuffer.
  • Adds rounded_selection setting to disable rounded corners for text selection.
  • Adds configurable timeout for MCP tool calls.
  • Adds custom API URL support for the DeepSeek provider.
+18 moreshow less
  • Adds \c and \C query pattern items in Vim/Helix mode to control case-sensitivity in buffer search.
  • Adds scrolling of the documentation panel alongside the completion menu via ctrl-d, ctrl-u, ctrl-e, and ctrl-y.
  • Adds a 'Wrap Selections in Tag' action for HTML, JSX, and similar languages.
  • Adds support for editing folders in WSL on Windows Alpha.
  • Improved font rendering via perceptual gamma/contrast correction, and improved text contrast when text is selected or highlighted.
  • OpenRouter provider now automatically retries on rate-limit or server errors.
  • Ollama provider now only sends the tools object to models that support it, preventing 400 errors.
  • Improved tool compatibility with OpenAI API-compatible providers.
  • Added LSP log tracing in remote collaboration and remote SSH environments.
  • Debugger now offers an option to show only stack frames from the user's project.
  • Extensions installation and update progress now shown in the activity indicator.
  • Markdown Preview now supports HTML <img src="..."> tags including SVG images.
  • Adds rewrap support in block comments.
  • Helix: x now respects count modifiers (e.g. 5 x) and pressing x on an empty line selects the current and next line.
  • Adds support for the Insert key from a character key location for keyboard layouts like neo2.
  • Adds ctrl-insert keybinding for copying in Markdown preview on Linux/Windows.
  • Python: virtual environment search now looks in intermediate directories between the worktree root and the pyproject.toml location.
  • Swapped default bindings for editor::NewlineBelow and editor::NewlineAbove on Linux and Windows to align with other editors.
└──▷ BREAKING ON UPGRADE
  • !Editing an old message in a native agent thread no longer automatically restores the checkpoint — it now only rejects unaccepted agent edits. Use the 'restore checkpoint' button to get the previous behavior.
Was this useful?

Google gemini-cli

Sources Release notes → v0.6.0-nightly.20250910.a31830a3 2 RELEASES · 2025-09-10 NOTES STABLE

An open-source AI agent that brings the power of Gemini directly into your terminal.

gemini-cli v0.6.0-nightly gains built-in docs, git, and PR review commands plus smarter tool-output summarization

└──▷ GET THIS VERSION
$ git clone --branch v0.6.0-nightly.20250910.a31830a3 https://github.com/google-gemini/gemini-cli.git
# already have the repo? check out this version:
$ git checkout v0.6.0-nightly.20250910.a31830a3
  • Adds new built-in commands for docs lookup, git operations, and PR review workflows.
  • Limits long tool output summarization to shell commands only, reducing unnecessary token usage for other tool types.
  • Moves IDE context filtering to the CLI side and updates port discovery logic for IDE integrations.
  • Adds a reusable EnumSelector UI component, enabling richer interactive selection prompts.
└──▷ BREAKING ON UPGRADE
  • !Org/repo install support has been removed; setups relying on installing extensions via org or repo references will break on upgrade.
1 more release in this issue · 2025-09-10
v0.5.0-preview NOTES STABLE

gemini-cli v0.5.0-preview adds docs/git/PR-review commands, runtime ripgrep download, and large tool-output file offloading.

└──▷ GET THIS VERSION
$ git clone --branch v0.5.0-preview https://github.com/google-gemini/gemini-cli.git
# already have the repo? check out this version:
$ git checkout v0.5.0-preview
└──▷ TRY IT
Install an extension by passing its source as a positional argument (new syntax after the flag was removed).
$ gemini extensions install <source>
  • Adds new /docs, /git, and PR-review slash commands for common developer workflows.
  • Downloads ripgrep at runtime (when enabled) instead of requiring a pre-installed binary.
  • Saves large shell-command tool outputs to a file and returns only truncated lines, preventing context overflow.
  • Improves discoverability of MCP slash commands in the UI.
  • Adds enum option support for settings fields, enabling constrained value selection.
+4 moreshow less
  • Allows typing in the chat input while MCP servers are still loading.
  • Makes source a positional argument in the extensions install command.
  • Supports hidden slash commands that opt out of autocompletion and /help discovery.
  • Moves scope settings to a new dedicated dialog.
└──▷ BREAKING ON UPGRADE
  • !The source argument for extensions install is now a positional argument instead of a flag; invocations using it as a named flag will break.
  • !Org/repo install support has been removed from extensions; setups relying on that install path will no longer work.
Was this useful?
Other / Uncategorized
◆  VECTOR DB RAG

LanceDB

Sources Release notes → python-v0.25.1-beta.0 2 RELEASES · 2025-09-10 NOTES STABLE

LanceDB python-v0.25.1-beta.0 adds mutual TLS (mTLS) support for remote database connections.

└──▷ GET THIS VERSION
$ git clone --branch python-v0.25.1-beta.0 https://github.com/lancedb/lancedb.git
# already have the repo? check out this version:
$ git checkout python-v0.25.1-beta.0
  • Supports mTLS (mutual TLS) authentication for remote database connections.
1 more release in this issue · 2025-09-10
v0.22.1-beta.0 NOTES STABLE

LanceDB v0.22.1-beta.0 adds mutual TLS (mTLS) support for remote database connections.

└──▷ GET THIS VERSION
$ git clone --branch v0.22.1-beta.0 https://github.com/lancedb/lancedb.git
# already have the repo? check out this version:
$ git checkout v0.22.1-beta.0
  • Adds mTLS (mutual TLS) support for remote database connections, enabling certificate-based client authentication.
Was this useful?

Weaviate

Sources Release notes → v1.31.14 NOTES

Weaviate v1.31.14 adds reasoningEffort and verbosity params to the generative-openai module.

└──▷ GET THIS VERSION
$ git clone --branch v1.31.14 https://github.com/weaviate/weaviate.git
# already have the repo? check out this version:
$ git checkout v1.31.14
  • Adds reasoningEffort and verbosity parameters to the generative-openai module for controlling OpenAI reasoning model behavior.
Was this useful?
◆  AI OBSERVABILITY

Langfuse

Sources Release notes → v3.109.0 NOTES

Langfuse v3.109.0 adds total cost column to dataset runs, score filters in dataset run items, and custom select filters in dashboard widgets.

└──▷ GET THIS VERSION
$ git clone --branch v3.109.0 https://github.com/langfuse/langfuse.git
# already have the repo? check out this version:
$ git checkout v3.109.0
  • Adds score filters to the dataset run items UI table, letting users filter run items by evaluation scores directly in the table view.
  • Adds a total cost column to the dataset runs table for at-a-glance cost tracking across runs.
  • Allows custom select filters in the dashboard widget flow, enabling more targeted metric views.
  • Pins select, annotation, and prompt metrics columns on horizontal and vertical scroll so key columns remain visible during table navigation.
  • Adds IP allowlist support for webhooks.
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 →