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.
Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explaining complex code, and handling git workflows - all through natural language commands.
Claude Code v2.1.235 adds optional spellcheck setting with aspell/hunspell/ispell support and hardens SendMessage size limits.
└──▷ GET THIS VERSION
$ git clone --branch v2.1.235 https://github.com/anthropics/claude-code.git
# already have the repo? check out this version:$ git checkout v2.1.235
└──▷ USE IT
Enable live spellchecking in the prompt input so typos are underlined as you type — requires aspell, hunspell, or ispell to be installed.
yaml
spellcheck: true
›Adds optional spellcheck setting that underlines misspelled words in the prompt input as you type, using your installed aspell, hunspell, or ispell.
›SendMessage now rejects messages too large for cross-session delivery up front instead of silently dropping them.
›Remote Control: claude rc now applies the same enterprise-gateway availability check as interactive startup.
›Improved memory and CPU usage while cloud sessions such as /ultrareview or /autofix-pr run in the background — their event streams are no longer re-scanned and re-rendered on every update.
›Context-limit error now indicates when auto-compact is off and points to /config to re-enable it.
+1 moreshow less
›Vim mode now preserves NORMAL mode and cursor position when toggling the detailed transcript (ctrl+o) or closing a panel.
Autonomous coding agent as an SDK, IDE extension, or CLI assistant.
Cline desktop beta adds realtime voice, model-driven image generation, cloud agent sessions, media playback, and SSH remote environments.
└──▷ GET THIS VERSION
$ git clone --branch desktop-beta https://github.com/cline/cline.git
# already have the repo? check out this version:$ git checkout desktop-beta
›Adds cloud agent sessions gated behind an explicit settings toggle and an onboarding panel for the GitHub connect flow, with model switching supported mid-session.
›Adds a media player in the desktop app capable of playing generated audio artifacts and displaying generated video.
›Adds realtime voice sessions with UI refinements for the desktop app.
›Adds an avatar/pet overlay to the desktop UI.
›Adds SSH remote environments as a proof-of-concept for connecting to remote development targets.
+3 moreshow less
›Adds animated reasoning and tool disclosures in the UI.
›Adds a beta release channel with in-app beta channel identity surface for the desktop app.
›CLI forwards generated images as ACP agent message chunks with type: image, data, and mimeType fields; non-image generated media (audio, video) is represented as a type: text placeholder chunk.
Cline desktop beta adds cloud sessions with cross-device sync, SSH remote support, voice input, and model-driven image generation.
└──▷ GET THIS VERSION
$ git clone --branch desktop-v0.0.14-beta.1 https://github.com/cline/cline.git
# already have the repo? check out this version:$ git checkout desktop-v0.0.14-beta.1
└──▷ HOW TO FIND IT
Enable cloud sessions to hand off a coding session between devices without losing context.
📍In the desktop app, go to Settings and enable the 'Cloud sessions' toggle. During onboarding, connect your GitHub account, then pick a repository and branch to start a cloud session.
›Adds cloud sessions (preview) — connect GitHub, pick a repository and branch, and run sessions in Cline's cloud with transcripts, approvals, and queued prompts synced across devices; enable via the Cloud sessions toggle in Settings.
›Supports renaming cloud sessions and switching models mid-session when using cloud sessions.
›Adds an SSH remote environment proof-of-concept for running sessions on remote machines.
›Adds an avatar overlay (preview): a floating desktop companion that reacts to active session state.
›Adds microphone voice input in the composer.
+4 moreshow less
›Adds model-driven image generation.
›Adds redesigned question prompts with animated reasoning and tool disclosures.
›Adds a GitHub integration step to onboarding.
›Installs side-by-side with the stable app and updates from its own beta channel, leaving stable installs unaffected.
Codex CLI alpha.23 adds exec-server forwarding mode, PowerShell Tree-sitter safety analysis, and hardens auth isolation for custom providers.
└──▷ GET THIS VERSION
$ git clone --branch rust-v0.148.0-alpha.23 https://github.com/openai/codex.git
# already have the repo? check out this version:$ git checkout rust-v0.148.0-alpha.23
›Adds exec-server forwarding mode via a new ExecServerSubcommand (accessible as codex exec-server <subcommand>), enabling the exec-server to be composed in additional topologies.
›Adds --remote <URL> flag to exec-server (now marked global) for registering the exec-server as a remote environment, and makes --strict-config global so subcommands inherit it.
›Adds a fail-closed Tree-sitter-based PowerShell lowerer that converts a conservative subset of literal PowerShell commands into argument vectors, rejecting dynamic expressions, #requires directives, and unrecognized syntax instead of guessing.
›Adds symlink-safe reading of sensitive files via read_sensitive_file_to_string, refusing to follow the final symlink component on Unix or reparse points on Windows.
›Rejects symbolic links used as memory workspace roots and removes links created during consolidation, preventing path-traversal out of the workspace.
+5 moreshow less
›Adds provider-owned authentication recovery, allowing providers to handle their own auth refresh flow independently.
›Scopes MCP resource reads by connector, preventing cross-connector data access.
›Requires fresh approval beneath denied permission paths, closing a re-approval gap in the permission model.
›Adds Guardian v2 approval review metrics, risk classification improvements, and Node REPL policy injection into Guardian review sessions.
›Reconnects WebRTC Realtime sideband transports on disconnection for improved reliability of real-time agent communication.
Codex CLI adds codex agents dashboard, desktop diagnostics in codex doctor, and queued-message dispatch for multi-process sessions.
└──▷ GET THIS VERSION
$ git clone --branch rust-v0.148.0-alpha.22 https://github.com/openai/codex.git
# already have the repo? check out this version:$ git checkout rust-v0.148.0-alpha.22
└──▷ TRY IT
Open the interactive agents dashboard to monitor and navigate running agent tasks in the TUI.
$ codex agents
›Adds codex agents as a dedicated interactive task dashboard command for monitoring and navigating agents in the TUI.
›Adds desktop app diagnostics and desktop update diagnostics to codex doctor, covering app-server handshake status, security enforcement (macOS Gatekeeper/XProtect, Windows), and update state.
›Adds a command to queue messages for existing sessions, enabling other processes to write to a running thread's durable queue and have idle threads dispatch them automatically.
›Adds configurable keyboard shortcuts for the agents dashboard.
›Restricts MCP HTTP redirects to the configured origin, tightening network policy for MCP app integrations.
+8 moreshow less
›Applies user MCP policy to selected executor plugins.
›Adds network policy metadata to environment configuration.
›Raises the GPT-5.6 maximum context window.
›Drops capabilities from Linux sandbox processes, hardening the sandbox security posture.
›Hardens Windows sandbox provisioning against reparse points.
›Surfaces interactive requests in realtime conversations.
›Requires approval for commands with dynamic shell words.
›Adds managed gates for in-app chat and dictation.
└──▷ BREAKING ON UPGRADE
!The experimental thread config endpoint has been removed (Remove the experimental thread config endpoint).
$ git clone --branch v2026.8.18 https://github.com/NousResearch/hermes-agent.git
# already have the repo? check out this version:$ git checkout v2026.8.18
└──▷ TRY IT
Update an existing Hermes install while getting accurate status if your branch is parked.
$ hermes update
›Adds hermes update parked-branch honesty, now accurately reporting when the local branch is parked rather than silently misreporting state.
›Adds NVIDIA SkillEvaluator Tier 1 advisory scanning on skill installs, running license and security checks before a skill is installed.
›Adds configurable timeout for cron media-send jobs, with support for manual-run attachments and surfacing of missed-fire events.
›Adds a tabbed SESSIONS|BOTS sidebar with per-bot hide/unhide controls.
›Adds desktop glass/translucency UI surface work including matte glass, a frost picker, and macOS pre-select behavior.
+2 moreshow less
›Adds native OS notifications for kanban board events.
›Extends Bot Mode group-chat support with fixes for long-running member turns, Markdown rendering, and cross-machine routing.
langchain-openai 1.5.2 extracts gateway metadata from response headers
└──▷ GET THIS VERSION
$ git clone --branch langchain-openai==1.5.2 https://github.com/langchain-ai/langchain.git
# already have the repo? check out this version:$ git checkout langchain-openai==1.5.2
›Extracts gateway metadata from response headers when available, surfacing routing and observability data from API gateway intermediaries.
langchain-openai 1.5.2a1 adds gateway metadata extraction, LangSmith gateway support, OpenAI 3.0 SDK, ChatOpenAICodex, explicit prompt caching, apply_patch tool, and more.
└──▷ GET THIS VERSION
$ git clone --branch langchain-openai==1.5.2a1 https://github.com/langchain-ai/langchain.git
# already have the repo? check out this version:$ git checkout langchain-openai==1.5.2a1
└──▷ USE IT
Set reasoning effort per-call to control how much reasoning an o-series model applies before responding.
python
from langchain_openai import ChatOpenAI
llm = ChatOpenAI(model="o3", reasoning_effort="high")
response = llm.invoke("Explain the halting problem.")
print(response.content)
Use the new ChatGPT Codex OAuth-backed model to run coding tasks via the Responses API.
python
from langchain_openai import ChatOpenAICodex
llm = ChatOpenAICodex()
response = llm.invoke("Write a Python function that parses JWT claims without a library.")
print(response.content)
Catch a context-window overflow explicitly so your app can truncate and retry rather than crash.
python
from langchain_core.errors import ContextOverflowError
from langchain_openai import ChatOpenAI
llm = ChatOpenAI(model="gpt-4o")
try:
response = llm.invoke(very_long_messages)
except ContextOverflowError as e:
print(f"Context exceeded: {e}. Truncating and retrying.")
›Adds reasoning_effort as a standard chat model parameter across OpenAI-compatible models.
›Supports LangSmith gateway through an environment variable (feat(anthropic,fireworks,openai): support langsmith gateway through env var).
›Extracts gateway metadata from response headers when available in ChatOpenAI.
›Adds ChatOpenAICodex OAuth-backed chat model for ChatGPT Codex interactions.
›Supports explicit prompt caching in ChatOpenAI.
+8 moreshow less
›Supports the apply_patch built-in tool in the OpenAI Responses API.
›Supports tool search in the OpenAI Responses API.
›Supports automatic server-side compaction for conversation management.
›Adds ContextOverflowError, raised in OpenAI and Anthropic integrations when context window is exceeded.
›Supports the OpenAI 3.0 SDK.
›Adds langchain-openrouter as a new provider package with streaming token usage support.
›Adds text_inputs and text_outputs fields to model profiles.
›Imputes placeholder filenames for OpenAI file inputs in core.
ONNX Runtime v1.28.1 adds device-free WebGPU compile-only sessions for offline model optimization without GPU hardware.
└──▷ GET THIS VERSION
$ git clone --branch v1.28.1 https://github.com/microsoft/onnxruntime.git
# already have the repo? check out this version:$ git checkout v1.28.1
›Adds device-free compile-only sessions in the WebGPU EP, enabling offline graph transformation and optimized-model serialization without access to GPU hardware.
An MCP that lets AI tools securely connect to your infrastructure, write IaaS code, debug issues, and assist during incidents - without risking production stability. Built for security teams to approve and infrastructure teams to experience like magic.
emisar v0.41.0 adds role-scoped consoles, live session reconnection on permission changes, and 100-pack catalog with new GCP, Consul, and HCP Terraform actions.
└──▷ GET THIS VERSION
$ git clone --branch v0.41.0 https://github.com/AndrewDryga/emisar.git
# already have the repo? check out this version:$ git checkout v0.41.0
›Role and scope changes (member role, runner access, pack access) now take effect immediately in open web sessions, reconnecting the session before stale authority can be exercised.
›Billing managers receive a finance-only console; operators can manage agents, approvals, and runbooks without gaining team or policy administration rights.
›Scoped admins are blocked from delegating more reach than they hold or triggering account-wide pack cleanup.
›Reads that expose every pack version or an exact run command now require a checked subject rather than relying on already-filtered callers.
›The read-only staff console now requires MFA proof tied to the current enrollment.
+7 moreshow less
›Console navigation, actions, filters, and empty states adapt to the member's role and access; billing managers no longer see a dead Dashboard link, and absent runner access is surfaced as a permission state.
›Restricted pack and action views now explain why results are limited.
›SSO group mappings now paginate in both directions.
›Pack catalog expands to 100 packs and 1,689 actions.
›Consul pack gains the ability to snapshot registration churn; the debugging pack adds bounded process context, environment-key, argv, and connection reads.
›GCP Monitoring pack adds Cloud Logging name and entry reads.
›HCP Terraform pack exposes replacement paths in plan summaries.
Composio OpenAI/Anthropic tool-call helpers now route execution through a Tool Router session, with error text preserved in { error } results.
└──▷ GET THIS VERSION
$ git clone --branch @composio/[email protected] https://github.com/ComposioHQ/composio.git
# already have the repo? check out this version:$ git checkout @composio/[email protected]
›OpenAI and Anthropic provider tool-call helpers can now execute through a supplied Tool Router session, with session meta-tools retaining their session context alongside existing user-ID direct execution.
›Anthropic helper failures now preserve error text in { error } results without altering successful payloads.
└──▷ BREAKING ON UPGRADE
!Custom provider subclasses overriding executeToolCall or handleToolCalls may require updates because these methods now accept session targets.
Composio @composio/[email protected] adds Tool Router session support for OpenAI/Anthropic helpers and extends SSRF guards to API-response URLs.
└──▷ GET THIS VERSION
$ git clone --branch @composio/[email protected] https://github.com/ComposioHQ/composio.git
# already have the repo? check out this version:$ git checkout @composio/[email protected]
›OpenAI and Anthropic provider tool-call helpers (executeToolCall / handleToolCalls) can now execute through a supplied Tool Router session, with session meta-tools retaining their session context across calls.
›Anthropic helper failures now preserve error text in { error } results instead of swallowing them.
└──▷ BREAKING ON UPGRADE
!Custom provider subclasses overriding executeToolCall or handleToolCalls may break on upgrade because these methods now accept session targets as an additional parameter.
Composio Anthropic helper routes tool calls through a Tool Router session and preserves error text in { error } results.
└──▷ GET THIS VERSION
$ git clone --branch @composio/[email protected] https://github.com/ComposioHQ/composio.git
# already have the repo? check out this version:$ git checkout @composio/[email protected]
›Enables the Anthropic provider tool-call helper to execute through a supplied Tool Router session, with session meta-tools retaining their session context alongside existing user-ID direct execution.
›Anthropic helper failures now preserve their error text in { error } results without altering successful payloads.
›Custom provider subclasses overriding executeToolCall or handleToolCalls now accept session targets, enabling session-aware routing in subclassed providers.
└──▷ BREAKING ON UPGRADE
!Custom provider subclasses overriding executeToolCall or handleToolCalls may require updates because these methods now accept session targets.
Weaviate v1.38.10 adds REST Search API endpoints for BM25, hybrid, near-object, and aggregate-counts queries, plus includeRoles in backup/restore and opt-in gRPC transport for GCS backups.
└──▷ GET THIS VERSION
$ git clone --branch v1.38.10 https://github.com/weaviate/weaviate.git
# already have the repo? check out this version:$ git checkout v1.38.10
└──▷ TRY IT
Back up a Weaviate collection and include RBAC role definitions so they are restored alongside the data.
THE AI TOOLCHAIN