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.
AutoGPT Platform adds Claude Opus 4.7, Redis Cluster, Web Push via VAPID, a Copilot Discord bot, and a redesigned Settings v2 UI.
└──▷ GET THIS VERSION
$ git clone --branch autogpt-platform-beta-v0.6.58 https://github.com/Significant-Gravitas/AutoGPT.git
# already have the repo? check out this version:$ git checkout autogpt-platform-beta-v0.6.58
›Adds SETTINGS_V2 feature flag to enable the new Settings v2 page layout, with dedicated sub-pages for API keys, integrations, profile, preferences, and creator dashboard.
›Adds ENABLE_PLATFORM_PAYMENT flag to gate the new subscription plan selection step and credit-grant paywall dialog during onboarding.
›Adds dynamic BlockCostType enum supporting SECOND, ITEMS, COST_USD, and TOKENS billing dimensions, with dynamic COST_USD billing and LaunchDarkly-configurable pricing tiers.
›Adds LaunchDarkly-configurable rate-limit multipliers with relative display in the UI.
›Adds a MAX subscription tier with LD-configurable pricing; unconfigured tiers are hidden automatically.
+9 moreshow less
›Adds Redis Cluster client support for multi-node Redis deployments.
›Adds Web Push notifications via VAPID for background delivery to users.
›Adds a Copilot-bot integration (Python / discord.py) for interacting with Copilot from Discord.
›Adds Claude Opus 4.7 model support.
›Adds inline picker-backed inputs via run_block, with support for AgentInputBlock subclasses.
›Adds message timestamps and accurate thought-duration timing display in Copilot.
›Adds a provider description SDK for the Settings v2 integrations page.
›Adds store creator filtering to show only creators with approved agents.
›Adds subscription credit grants, paywall gate, dialog UX, and cross-pod cache for subscription management.
langchain-perplexity 1.2.0 adds PerplexityEmbeddings class and overhauls the Perplexity integration with the official SDK and Search API.
└──▷ GET THIS VERSION
$ git clone --branch langchain-perplexity==1.2.0 https://github.com/langchain-ai/langchain.git
# already have the repo? check out this version:$ git checkout langchain-perplexity==1.2.0
└──▷ USE IT
Generate embeddings for a batch of texts using the new Perplexity embeddings model.
python
from langchain_perplexity import PerplexityEmbeddings
embeddings = PerplexityEmbeddings()
vectors = embeddings.embed_documents(["What is zero-day exploitation?", "Explain lateral movement."])
print(vectors[0][:5])
›Adds PerplexityEmbeddings class for generating embeddings via the Perplexity API.
›Overhauls the Perplexity integration to use the official Perplexity SDK and Search API.
LangChain 1.2.16 adds content-block-centric streaming and agent-type tagging on create_agent calls.
└──▷ GET THIS VERSION
$ git clone --branch langchain==1.2.16 https://github.com/langchain-ai/langchain.git
# already have the repo? check out this version:$ git checkout langchain==1.2.16
›Adds ls_agent_type tag automatically on create_agent calls for improved agent observability and tracing.
›Introduces content-block-centric streaming (v2) in core for finer-grained streaming of LLM responses.
›Adds a benchmark command for measuring LangChain initialization and middleware performance.
LangGraph checkpoint 4.1.0a1 adds timer support (alpha) and a new DeltaChannel for efficient checkpoint reconstruction.
└──▷ GET THIS VERSION
$ git clone --branch checkpoint==4.1.0a1 https://github.com/langchain-ai/langgraph.git
# already have the repo? check out this version:$ git checkout checkpoint==4.1.0a1
›Adds alpha timer support for checkpoint-based workflows.
›Introduces DeltaChannel: stores sentinel values in blobs and reconstructs state from checkpoint_writes rather than full snapshots.
Enable GPU passthrough so a sandboxed agent can run local GPU workloads inside Docker.
yaml
# In your OpenClaw config:
sandbox:
docker:
gpus: all
›Adds /codex computer-use status/install commands plus marketplace discovery, optional auto-install, and fail-closed MCP server checks before Codex-mode desktop-control turns start.
›Adds bundled DeepInfra provider with DEEPINFRA_API_KEY onboarding, dynamic OpenAI-compatible model discovery, image generation/editing, image/audio media understanding, TTS, text-to-video, and memory embeddings.
›Adds Tencent Yuanbao channel plugin (openclaw-plugin-yuanbao) with WebSocket bot DMs, group chats, and a new docs/channels/yuanbao.md quick-start guide.
›Adds full QQBot group chat support: history tracking, @-mention gating, activation modes, per-group config, FIFO message queue with deliver debounce, C2C streaming via StreamingController, and chunked sendMedia for large files.
›Adds opt-in operator-managed outbound proxy routing via proxy.enabled + proxy.proxyUrl / OPENCLAW_PROXY_URL with strict http:// forward-proxy validation and loopback-only Gateway bypass.
+9 moreshow less
›Adds opt-in sandbox.docker.gpus passthrough so local GPU workloads can run inside Docker sandbox containers when the host runtime supports --gpus.
›Adds node.presence.alive authenticated protocol event and node.list last-seen fields so background iOS and Android wakes mark paired nodes recently alive without treating them as connected.
›Accepts non-image attachments through chat.send by staging them as agent-readable media paths instead of silently dropping files.
›Adds streaming.preview.toolProgress: false config flag to hide interim tool lines in live Matrix preview edits while keeping answer previews.
›Adds generic host hooks in Plugin SDK for session state, next-turn context, trusted tool policy, UI descriptors, events, scheduler cleanup, and run-scoped plugin context.
›Exposes shared channel route normalization, parser-driven target resolution, raw-target compact keys, parsed-target types, and route comparison helpers through openclaw/plugin-sdk/channel-route.
›Wires manifest modelCatalog.aliases and modelCatalog.suppressions into model-catalog planning and built-in model suppression, enabling provider rows/aliases/suppressions to be declared in plugin manifests.
›Adds plugin compatibility warnings for deprecated implicit startup loading so plugin authors can migrate to explicit activation.onStartup metadata.
›Attaches versioned structured approval metadata to pending Matrix approval messages so capable clients can render richer approval UI.
PydanticAI v1.88.0 adds output validate/process hooks, cross-provider service_tier, Anthropic fast mode, and new UI sanitization APIs.
└──▷ GET THIS VERSION
$ git clone --branch v1.88.0 https://github.com/pydantic/pydantic-ai.git
# already have the repo? check out this version:$ git checkout v1.88.0
└──▷ USE IT
Enforce consistent service priority across providers without per-model configuration.
python
from pydantic_ai import Agent
agent = Agent(
'anthropic:claude-opus-4-6',
model_settings={'service_tier': 'priority'},
)
result = agent.run_sync('Summarize this document.')
print(result.output)
›Adds prepare_output_tools hook alongside prepare_tools — prepare_tools is now scoped to function tools only, while output validate/process hooks give fine-grained control over output tool execution.
›Adds cross-provider service_tier model setting with support for Anthropic, Gemini API, and Vertex Priority PayGo.
›Adds fast speed mode for Anthropic Opus 4.6.
›Adds UIAdapter.sanitize_messages and allowed_file_url_schemes to the UI adapter for controlling which file URL schemes are permitted in messages.
›Supports OpenAI Responses phase field on assistant messages.
└──▷ BREAKING ON UPGRADE
!prepare_tools is now scoped to function tools only; callers relying on it to prepare output tools must migrate to the new prepare_output_tools hook.
Semantic Kernel .NET 1.75.0 adds Vertex AI embedding dimensions support and read-only vector store property interfaces.
└──▷ GET THIS VERSION
$ git clone --branch dotnet-1.75.0 https://github.com/microsoft/semantic-kernel.git
# already have the repo? check out this version:$ git checkout dotnet-1.75.0
›Adds dimensions parameter support to Vertex AI embedding services in .NET, enabling control over output embedding size.
›Introduces read-only property interfaces on CollectionModel/builder, giving callers a stable, immutable view of vector store collection metadata.
›Updates .NET SQL Server vector search to use the latest VECTOR_SEARCH() syntax.
›Applies FunctionChoiceBehavior filters to OpenAI responses agent tools in Python, so function selection constraints are now respected in agent workflows.
$ git clone --branch v1.0.0 https://github.com/zed-industries/zed.git
# already have the repo? check out this version:$ git checkout v1.0.0
└──▷ HOW TO FIND IT
Jump to any commit by ref without leaving the editor — useful for quickly reviewing a tagged release or a specific SHA during an investigation.
📍git: view commit → type HEAD~5 (or any SHA, branch, or tag) in the prompt
Navigate a large file hands-free during a code review by toggling bookmarks on key lines and cycling through them.
📍editor: toggle bookmark # mark the current line
editor: go to next bookmark # jump forward
editor: go to previous bookmark # jump back
editor: view bookmarks # see all bookmarks in a panel
›Adds editor: toggle bookmark, editor: go to next bookmark, editor: go to previous bookmark, editor: view bookmarks, and workspace: clear bookmarks command palette actions; bookmarks appear in the gutter and persist across sessions.
›Adds git: view commit command palette action to open any commit by git ref (HEAD, SHA, branch, or tag) in the commit view.
›Adds interleaved_reasoning option to OpenAI-compatible model configurations.
›Adds DeepSeek-V4-Pro and DeepSeek-V4-Flash models to the AI agent.
›Loads the commit message template from the file set in git config commit.template when starting a commit in the git panel.
+13 moreshow less
›Adds auto-completion of action names in the command_aliases setting.
›Adds 'Copy Link' to the right-click context menu on links in Markdown views (agent panel and Markdown preview).
›Adds a 'New Center Terminal' button to the window context menu.
›Adds GIF animation support in the Markdown Preview.
›Adds 'Configure Provider' menu item to Copilot and Codestral providers.
›Adds select-all and deselect-all checkboxes to the git panel.
›Associates .mlx file extensions with OCaml.
›Adds an action to copy the list of installed extensions to the clipboard.
›Enables diagnostic underlining in split editor views.
›Reuses existing SSH ControlMaster sessions for remote development instead of prompting for credentials again.
›Switches all OpenAI model requests to always use the Responses API.
›Improves fuzzy matching in the command palette, branch picker, tab switcher, and recent projects picker to support multi-word queries.
›Shows editor: format selections in the command palette only when the active formatter supports range formatting (e.g. Prettier or LSPs like clangd that advertise the capability).
llama.cpp b8980 adds configurable vmem size and buffer counts for Hexagon/Snapdragon backends.
└──▷ GET THIS VERSION
$ git clone --branch b8980 https://github.com/ggml-org/llama.cpp.git
# already have the repo? check out this version:$ git checkout b8980
›Adds host-side control over maximum vmem size for the Hexagon backend, with a new default of 3.2 GB and support for autodetection of available vmem space.
›Increases Hexagon op buffer count to 16 (matching the maximum number of mmaps) and moves pinned mmap management to the host.
Phoenix client v2.6.0 adds a session notes API endpoint, dataset upsert support, and extended prompt invocation parameter types.
└──▷ GET THIS VERSION
$ git clone --branch arize-phoenix-client-v2.6.0 https://github.com/Arize-ai/phoenix.git
# already have the repo? check out this version:$ git checkout arize-phoenix-client-v2.6.0
›Adds a session notes API endpoint (add session notes endpoint) for annotating sessions via the API.
›Introduces dataset upsert support, enabling create-or-update semantics when uploading datasets.
›Gates v15 dataset upload parameters by server version, ensuring compatibility across Phoenix server versions.
›Extends database prompt types to support additional invocation parameters.
└──▷ BREAKING ON UPGRADE
!Dataset upsert changes the behavior of dataset uploads; existing workflows relying on the prior insert-only semantics may be affected on upgrade.
Phoenix v15.0.0 adds dataset upsert support and extended DB prompt types for invocation parameters.
└──▷ GET THIS VERSION
$ git clone --branch arize-phoenix-v15.0.0 https://github.com/Arize-ai/phoenix.git
# already have the repo? check out this version:$ git checkout arize-phoenix-v15.0.0
›Adds dataset upsert capability, enabling insert-or-update semantics when pushing records to a dataset.
›Extends database prompt types to support invocation parameters, broadening the range of prompt configurations that can be stored.
Phoenix v14.17.0 adds custom provider support for agents, a session notes API endpoint, and a JSON schema for settings config.
└──▷ GET THIS VERSION
$ git clone --branch arize-phoenix-v14.17.0 https://github.com/Arize-ai/phoenix.git
# already have the repo? check out this version:$ git checkout arize-phoenix-v14.17.0
›Adds a session notes API endpoint for programmatic annotation of sessions.
›Adds JSON schema for the settings config, enabling editor validation of Phoenix configuration files.
›Agents now support custom providers and secret store values for extended LLM integrations.
THE AI TOOLCHAIN