IT
Agent frameworks, SDKs and CLIs, open source releases, local inference and coding tools.

10 security checks for AI-built apps before sign-up
Testing whether an app’s features work and whether it can safely handle real user data are separate jobs.
Last 7 daysNov 17 – Nov 22
OpenAI CLI makes automatic updates opt-in in the new release
Future updates install automatically only if the feature is enabled, and the release notes leave open whether existing installations keep their settings.codex-acp standard and preview entries move to version 2.9.0
The standard and preview entries previously listed different versions in the Agent Client Protocol registry.
DEV publishes a beginner’s guide to Craft Anything AI
The guide explains how to choose a website, app or game to rebuild and how to get started at no cost.
LangGraph 1.2.24 fixes message clearing and checkpoint bugs
The release locates the fixes in the message reducer and DeltaChannel.
Kimi Code CLI supports AWS credentials for Amazon Bedrock
The Bedrock connector generates a short-lived API key from existing AWS credentials when no separate key is configured.AWS adds shared cache skill for self-hosted language models
KVTrace estimates the benefits of a shared cache before deployment.Tripo AI unveils 3D tools for game developers at G-Star 2026
Tripo AI’s tools turn images into 3D models and help game developers carry the resulting assets into later stages of production.
AI illustrationHugging Face unifies generation cache names in Transformers
Support for the old cache argument names ends at different versions for `forward` and `generate`.CodeRabbit Verify tests code in running apps in public beta
Verify turns code changes into test scenarios and keeps records developers can use to investigate failures.
AI illustrationTeam Ninja sees AI as a tool for more efficient development
Team Ninja’s producers say AI can make development more efficient and people decide what makes the action fun.
AI illustrationPraxi-Bot sets out a design for approving model tool actions
Once a shell and browser are connected, desktop agent developers still have to decide which actions proposed by the model can run.
A developer details ElevenLabs voice output in Electron apps
A DEV Community post uses lessons and to-do lists to show where desktop apps can use voice output.
AI illustrationCodex makes its terminal workspace a default part of the CLI
Developers can search document previews and view images and subagent transcripts alongside the main conversation.
AI illustrationSwiftPY makes asynchronous processing the default for Python
SwiftPY makes asynchronous processing the default and targets developers familiar with Laravel.
Basic International unveils AI SyncOps Studio for businesses
Basic International’s AI centre uses design specifications to divide work among AI agents.
Cursor and Windsurf price comparison links to wrong company
Before using a shared price in a team budget, check the product name on the linked pricing page.
Google rolls out Gemini 4 Neon to developers on its paid API
Input, reasoning and output share a token limit, and the API rejects out-of-range sampling settings.Australis puts plan approval into app creation
A workflow linking English requests to planning, building and checking offers a basis for judging what follows code generation.
Google publishes website access guidance for Google-Agent
The help document covers access for Google-hosted agents that browse websites and perform tasks at users’ request.Gamehaus cuts AI game prototyping labour by about 90%
The company based its internal estimate on developer working days and said it also automated SDK integration and analytics event tracking.
browser-use makes 15-times token pricing the default
Developers must explicitly pass llm='bu-fast' to retain the previous model and pricing.
InfoQ opens five-week cohorts for senior practitioners
Participants test approaches in their own work and compare results with peers.
Qwen Code unifies terminal and subagent execution
A shared agent loop becomes the default, while reasoning effort choices are limited to levels supported by the active model.
Free product-photo tool joins toolbox for Chinese sellers
The developer described getting the model to understand Chinese product descriptions as harder than generating images.
Agent memory test highlights risks from stored secrets
A fake-key test raises two linked issues: removing secrets before storage and their later reuse through agent memory.
Free AI coding tool leaves model costs unclear
Assessing a free coding tool as an alternative to paid software means separating model costs from the scope of support.Google introduces AdMob agent skills for ad coding
AdMob’s coding instructions are available now, while the server connecting developer environments to account data remains in early testing.Hugging Face routes language models through OpenRouter
Failed requests are retried with another provider, while support for users' own provider keys ends for language models.IBK builds AI coding agent on internal infrastructure
The Korean bank plans to widen staff participation in improving work processes through a development tool built for financial-sector security constraints.Brave Search API allows agent use but limits result storage
Monthly credits are separate from a free plan, while permission to use the API with agents does not extend to storing search results.Nota supplies models and servers for Dube's coding agent
The supply covers model optimisation and server configuration for a coding agent designed to run on financial firms' closed networks.1 week agoNov 10 – Nov 16
Databricks SQL commands target coding agents
A DEV article examines how coding agents query Databricks SQL when table data and the command-line tool are already available.
AI-written code leaves the burden of software design
A dev.to summary described how deciding a product's direction took time even when AI wrote the code quickly.
Postman drill examines where agent-handled tokens end up
A staged drill using a fake token shifts attention from whether an agent completes a replacement to where secrets remain.
AWS details conditions for LLM gateway budget limits
Model pricing must be configured for spending limits to work, while developers need LiteLLM Enterprise to issue their own keys.
Wapzen targets AI assistance on existing WhatsApp numbers
A developer describes Wapzen as an AI agent for small businesses that want to keep the number their customers already know.
ElevenLabs voice guide targets Next.js developers
A DEV guide presented voice assistants and accessibility as uses for web app speech, while its service quality claims reflected the author's assessment.
Deleted checks matter more than an AI coding agent's summary
The merge scenario in a DEV post turns on what a deleted check protected, rather than how much the agent wrote.
Codex command rules give blocking priority over permission
A DEV.to post explained which command rules take priority when they overlap and how to test them before use.
HIVE402 rebuilds homepage around buyer questions
The redesign shifts the focus from a service catalogue to who takes on a job, what it costs and what the work delivers.
Keychron outlines magnetic keyboard firmware plans
The MKF introduction page sets out Keychron’s aims, while public code contributions remain a separate question.
AI image storage links originals to time-limited access
A DEV Community post recommends checking stored image details after uploading the original file, then providing access through links that expire.
AI app deployment checks turn to payment conditions
A feature working in a demo and an app handling payments without human oversight are different tests.
AI test answers can vary with the same code and prompt
A single AI test response gives quality teams little basis for judging consistency across repeated runs.
Hexaphase proposes keeping AI project memory in repositories
The Hexaphase proposal puts continuity between sessions at the centre of AI development workflows.
HyperS2C launches two AI agents for verification
The agents form part of a wider toolchain and platform launch, whose product count does not measure community response.
Arui.AI puts multilingual support at the start
Arui.AI's launch put language support alongside its creative tools from the outset.
Tiro offers dictation and translation within 0.5 seconds
Tiro's official blog lists API access covering notes, folders, audio files, glossaries, wikis and organisations.
Free AI code review tools differ in access and limits
A DEV post compares free tools for public repositories by application requirements, usage limits and whether maintainers must run them themselves.
dev-diagnosis examines coding agents' unnecessary reading
The tool examines repository reading scope and token costs, while adoption requires a separate assessment of availability and commercial terms.
Hugging Face unifies model export return format
The update brings exported models under a shared interface for sharing and execution, while changing how developers save them.
How repeated token choices shape language model responses
A DEV.to explainer uses the token selection loop to explain generation settings and the limits of prompt instructions.
AgiBot releases LIFE for robot sensing and memory
LIFE links continuous sensing with memory reanalysis, while its performance claims rest on AgiBot's own tests.
AWS opens web search API for direct calls without a gateway
For game development pipelines that call search directly, developers manage retries, caching and error handling.
StareBrain details screen reading and action dispatch
The development series moves from user input and plan validation to how the Android agent reads screens and sends interface actions.
Guarapay’s AI card raises spending authority questions
Guarapay’s freeze, approve, refreeze proposal separates an agent’s ability to execute payments from its permission to spend.
Whisper setup guide presents local text prompts
A post titled as a Whisper speech recognition guide presents local model setup and a text question example in its summary.
Codex sends tool updates incrementally by default
Developers can turn off the default behaviour by setting a specified value in the Codex configuration file.
Copilot CLI adds regex rules for enterprise shell access
The update expands command permission rules, fixes the order of follow-up instructions and preserves saved settings when configuration files cannot be read.
JARVIS Chat launch plan centres on AI integration
The DEV post presented bringing several AI capabilities together as JARVIS Chat’s distinguishing feature.
Lotte Innovate links AI monitoring expansion to web security
The Elec reported a plan to connect web protection to Lotte Innovate’s own security monitoring platform and extend integrated services to external businesses.
NVIDIA adds faster double-precision matrix calculations
NVIDIA says new cuBLAS algorithms accelerate matrix multiplication on integer Tensor Cores while preserving double-precision accuracy.
Parallel Decisions Fast joins AI Gateway free for now
Answer probabilities can guide automated actions, but cited evidence applies to the whole request rather than individual answers.
More game developers use generative AI at work
Developers’ use of generative AI at work must be read separately from the prevalence of generation during gameplay.
LaunchVet maker revisits fixes to its demo app
The recheck and free core offer need to be read as the tool maker's own assessment.
n8n Labs releases Nod personal agent as a research preview
Nod is presented as a personal agent with its own computer and browser, but adoption depends on its terms of use and where it runs.
Google releases AQuA to diagnose AI agent failures
AQuA links recurring failures to deployed code, with session replay used to check prompt changes.
OpenAI releases Codex 0.167.1 on GitHub
The official release links to a comparison with the previous version without listing specific fixes.
Windsurf switching decisions start with the coding task
Comparing code writing, refactoring and debugging starts with deciding which task a developer wants to improve.
NC AI shares game-making methods at first builders meetup
Sessions covered using VARCO-generated assets in games, connecting AI tools and managing output quality.
Vercel expands AI Gateway generation and decision models
Document generation and decision outputs serve different tasks, while decision calls count towards the same budgets as other model requests.
AI Game Lab Day showcases game development tools
Engine tools and command-line workflows featured alongside developer presentations and individual consultations for resident companies.
Litmus sets out a simpler approach to prompt-change checks
Litmus's case for simpler checks explains why it was built, but does not establish how well it detects errors.
AI drafts Freshdesk replies as private notes
A DEV Community post outlines a workflow that leaves reply drafts for support staff to review and routes suspected defects to GitHub issues.
agent-browser targets checks on AI-generated pages
A DEV Community article presents browser interaction and result checking after code generation as the tool's purpose.
ByteDance releases lower-cost Seedance 2.5 Flash
The reported per-second price applies at 720p, while API access has a separate opening date.
Block offers AI agents as teammates for businesses
Offering internal tools to businesses makes it necessary to distinguish agents’ collaborative roles from users’ rights to use them.
AI model switches take weeks as apps need reworking
Easier connection changes still leave businesses with prompts, tools and evaluation systems to adjust.
Spider Networks unveils AI agent rules hub
KUKURI lets users select rules for their work instead of writing complex AI instructions from scratch.
ElevenLabs releases Scribe v3 transcription models
Neural Brew reported a company-published reduction in transcription errors and support for real-time speaker identification.
Google launches Hydrogen for faster coding and agent tasks
The reported task-speed gain uses an earlier lightweight model as its baseline, while the overall performance comparison uses a higher-tier model.
Codex adds terminal chat tabs and restart recovery
Users can reopen saved side chats and search older tasks by name, prompt preview or working directory.
AI memory test checks what survives a correction
An evaluation proposal checks whether correcting one fact preserves other memories and carries through to a new conversation.
AWS introduces UAEF to unify AI agent evaluation
UAEF converts execution records from different frameworks into a common format so teams can compare results using shared metrics.
Codex works directly with Home Assistant settings
The specific configuration work provides the basis for assessing one user's account of changed coding habits.
AWS outlines three Codex access paths through Bedrock
AWS set out how teams can choose between direct access using existing credentials and gateways that enforce controls on model requests.
Replit users bear responsibility for Agent-generated code
A security checklist described Replit’s division of responsibility between users who own generated code and the company that protects the platform.
Design proposes fixed priorities for AI code patch approval
The proposal puts rules for resolving conflicting validation signals at the centre of patch approval.
Meshy DCC Core fixes agent disconnections in `0.2.4`
The update prevents retries when a request's result is unknown and checks app readiness through the path used to send requests.
llama.cpp changes server tool-call controls
The release changes server tool controls and adds chat search, while its notes omit the licence and minimum VRAM requirements.
QCon AI New York programme covers agents and code review
Engineers’ accounts of running AI systems frame a programme covering agent permissions, code review and team adoption.
Coding AI search results vary by task across six methods
Concrete tasks and ambiguous requirements favoured different code search methods, making task type a starting point for game developers choosing coding AI tools.
MoveBench joins public API benchmark tally
Benchmark counts and leaderboard entry counts measure different changes in the evaluation list.
Anthropic urges review of prompts in Fable 5.2 guide
The guide addresses development problems including early stops during long tasks and the scope of code changes.
TTSE separates agent experience into facts and tips
TTSE proposes a way to classify agent experience, while reported gains remain the development team's own results.
Web update breaks nightly tests for an unchanged agent
Moved buttons and renamed fields raise compatibility concerns for evaluations that rely on reading and operating a screen.
DeepL unveils Aure to track changes in spoken expression
DeepL described a move from preserving a speaker’s voice to tracking changes in delivery, but its Korean product page still labels voice-to-voice translation as coming soon.
qwen-code releases TypeScript SDK v0.1.32
The release notes list different versions of the bundled command-line tool.
Lotte Innovate plans AI security service with Thales tech
Lotte Innovate plans to connect web protection to its monitoring platform and extend integrated security services beyond Lotte affiliates.
AI leaderboard separates accuracy from self-error detection
The evaluation distinguishes getting an answer right from recognising an error in the model's own answer.
Qwen Code preview removes separate Node.js install
Packages for major operating systems support session resumption, while preview interfaces, settings and behaviour remain subject to change.
AWS adds Kimi K3.1 Pro Highspeed to Amazon Bedrock
Existing API calls remain compatible, while Flex pricing cuts costs in exchange for potentially longer response times.
Hugging Face proposes a shared decision-model interface
The proposed interface brings probabilities for predefined answers and different question types under a common input and output contract.
Developer automates app language additions and deployment
The setup checks a language queue daily and runs a coding agent every three days, linking language additions to deployment, checks and email reports.
DeepSeek Harness reaches 1.54 million daily active users
Company-reported usage supports DeepSeek's entry into the market, but does not establish how its product performs against rivals.
A closer look at engineering teams before sales outreach
The proposed sales approach separates a company's industry, headcount and funding fit from its likelihood of buying.
Offline voice apps separate downloads from model use
Model files, companion files and runtime use represent different states that an offline AI setting needs to distinguish.
TwoLineCloud launches MuSAnga for system modernisation
According to ZDNet Korea, TwoLineCloud said MuSAnga analyses system structures and data relationships to propose cloud migration priorities and costs.
Hive Axil targets formal launch next year
The AI-focused development kit targets service functions such as login and payments that AI agents need to understand and use.
Claude Fable 5.2 divides development work among AI agents
The reported speed gain compares the new models with version 5.1, while Fable adds request analysis and task allocation for complex development work.
Decision models reshape AI Gateway requests and spending
Decision models account for sharply different shares of requests and spending, while new teams’ provider choices and spending shares also diverge.
AWS details seven-layer reference architecture for AI agents
The framework lets teams swap the technology within each layer, but the announcement does not name a licence for its sample repository.
qwen-code v0.27.1 adds opt-in Mem0 memory for managed agents
The CLI also warns of Wasm memory spikes on Node 24 and 25, while the release notes give no commercial or redistribution terms
qwen-code TypeScript SDK v0.1.30 includes CLI version 0.27.0
The release notes give only the bundled version and its build origin, so the changes have to be traced in the commit history.
AWS adds always-on Kiro Bots to Kiro Crew to organise agents
AWS points to a lighter load of agent supervision, but no measurement shows how far that load falls.
Meta open-sources Astryx, a design system made for AI agents
The system reportedly aims to reduce the inconsistencies that arise when AI composes screens arbitrarily, but the report omits the licence terms outside development teams need to assess.
ETRI makes its no-code AI development tool TANGO open source
TANGO lets users select functions in a web browser to develop and run AI models, but its licence terms and adoption cases are not yet confirmed.
Qwen Code SDK 0.1.29 reaches GitHub after a workflow failure
Fixes to long-session caching and a memory setting sit under bundled CLI entries, where the SDK number alone does not show them.
Qwen Code TypeScript SDK v0.1.31 ships bundled CLI at 0.27.1
The release notes list bundled CLI versions 0.27.1 and 0.27.0 and point developers to the commit history for changes.
2 weeks agoNov 03 – Nov 09
Aircheck gathers all posts agents queued today on one screen
The title of the Show HN post does not give a licence or say whether the tool runs locally.
Google ships Agent Substrate 1.0 on GKE and gives it to CNCF
Agent Substrate moves to the CNCF alongside gVisor and OpenRL, but public documents give no test conditions for its density figures and no list of breaking changes.
AWS adds Gateway policy layer to Okta agent delegation setup
Per-user permission decisions stay on the tool servers, and the AWS post gives no measurements at the scale of hundreds of agents.
Taxbit details tax AI design that keeps identity from models
The assistant for a regulated sector keeps permission scope outside the model, and Taxbit did not name the model that runs inference.
Hugging Face's transformers v5.22.2 hardens Jinja sandboxing
The patch hardens Jinja sandboxing, but the release notes do not name the specific flaw it was meant to block.
OpenAI Codex 0.167.0 gains backend switching in the terminal
Saved chats stay visible after a model provider change, and system proxy support arrives as an experimental setting.
Copilot CLI v1.0.107 refuses device code login on HTTP hosts
Plugin rules marked `alwaysApply: false` now reach the prompt only when a user turns them on.
NVIDIA publishes DSX OS procedures for managing GPU clusters
Three open source tools split configuration, certification and operation, but the post gives no release number or general-release schedule for DSX OS as a whole.
Amazon Quick lets a separate model approve its agent actions
Smart Approvals is on by default, so AWS advises administrators to check first which actions stay under human approval.
U Mobile runs its AiDaH helpdesk on Amazon Bedrock AgentCore
AWS says U Mobile's helpdesk settles about 70% of requests without a model call, but the passage reporting a ticket cut of up to 84.5% gives no measurement period.
Claude Code v2.1.318 prints login expiry on the command line
Self-hosted runner log lines have changed, so scripts that search for specific strings need updating.
Gen Digital details Norton Revamp's news engine built on AWS
An AWS blog post sets out how the engine receives search results inside AWS and offers operating advice, but it does not give the cost of search calls.
IBM's 20B slide model Palette arrives with no stated licence
IBM researchers report second place on the three-domain PresentBench mean, but their blog post does not say where the weights are distributed or what commercial terms apply.
NVIDIA proposes AICR for Kubernetes setup on Vera Rubin GPUs
NVIDIA says Kubernetes configuration for new GPU hardware is a burden for operators because every layer, from drivers to schedulers, must fit together.
3 weeks agoOct 26 – Nov 02
Python 3.15 dependency errors bring advice to delete uv.lock
uv's documentation tells projects to keep the lock file in version control, while the advice to delete it comes from personal experience.
ParticleDB ties its general release to an independent retest
Orkes says all identified issues are addressed, while ParticleDB's release status document separates in-house test runs from independent audits.
Claude Code retires Opus 4.8 fast mode with no Opus fallback
After removal on 30 Nov 2026, fast mode requests on Opus 4.8 return an error instead of moving to standard Opus 4.8.
NVIDIA post ties NemoClaw agents to policy-enforcing runtime
NVIDIA presents a policy-enforcing runtime as a requirement for more secure agent deployment, but the post's introduction gives no version number or licence terms.
NVIDIA publishes guide to checking GPUDirect Storage set-ups
The post covers the data path between GPUs and storage, but its introduction gives no supported versions or performance figures.
transformers patch v5.21.1 fixes remote code loading in vllm
The release notes do not name the version where the error began, and `v5.22.1`, published the next day, changed only documentation.
Mistral publishes Large 4 weights under modified MIT licence
Commercial use is allowed, but a model-as-a-service business needs a separate commercial licence.
Liquid AI's LFM3 models run on Linux and Windows with Ollama
The models now run beyond Apple silicon, but the release notes do not give their licence or minimum VRAM.
Claude Code cuts system prompts for unrecognised model names
Bedrock, Vertex, Foundry and Mantle users can keep the full prompt with an environment variable.
NVIDIA ties Ling-3.0-flash bottleneck on DGX Spark to memory
NVIDIA said reading model weights drove execution time in a batch-size-1 evaluation of the Inclusion AI model.
Transformers v5.21.0 forces changes to custom parallel plans
Custom tensor-parallel and mixture-of-experts plans face changes, while the release notes give no licence or commercial terms for the new models.
NVIDIA pitches Minerva to maximise GPU yield in AI factories
NVIDIA aims Minerva at AI factory operations split across many teams, but the opening of its blog post gives no release number or licence terms.
Cline desktop app gains a diff review pane for changed files
A terminal panel and local model settings arrive, and default models change for 56 providers.
Cline SDK v0.2.0 reconnects restarts to sessions in progress
The model request retry limit falls from 5 to 3, and a start request carrying a prompt is rejected while a run is in progress.
AWS shows a semantic layer for AI agents inside EDB Postgres
The same agent code returns different results depending on the database role that runs it.
Oct 2026Oct 12 – Oct 25
Cursor Origin API adds allowlist option for app-declared IPs
Approval counting and auto-merge rules also changed, so merge requests from existing integrations can be refused.
Ollama v0.40.11 brings macOS speech generation and d3 models
The release notes state no licence and give no VRAM requirements for each d3 size
TensorRT 11.1 adds automatic tuning customised to each model
NVIDIA frames neural network deployment as a compilation problem with several valid answers and automates optimisation for each model.
Codex 0.164.0 changes command input and authentication flows
The release notes list no breaking changes, but resumed sessions now restore different settings.
Microsoft posts ThinkingBox-Training agent RL tool on GitHub
Company figures show more tasks passed on every attempt and fewer solved at least once, and the licence name is not confirmed.
NVIDIA publishes guide for Python data reads into GPU memory
NVIDIA points to the efficiency of moving data from storage as compute and datasets grow together.
Claude Code v2.1.309 fixes sandbox and permission check gaps
A flaw that ran shell startup files outside the sandbox is fixed, and the release also changes behaviour that affects automation scripts.
Agents SDK adds Pi 1.4 support but deployments cannot revert
A session storage schema change blocks any return to Pi 1.3, and the announcement does not say how long the migration takes.
NVIDIA describes RFD4-Proteina method for designing proteins
A developer blog post frames protein design as a search problem and covers reward-guided inference-time search.
NVIDIA says AI agent controls must operate outside the agent
NVIDIA said monitoring and enforcement need to operate beyond an AI agent's reach because an agent that completes its task can still cross a security boundary.
Cursor makes its Agent SDK generally available to developers
Bots defined as code folders can be evaluated and deployed to Cursor or a user's own infrastructure, but the changelog omits the licence and deployment pricing.
transformers v5.20.2 fixes input type flaw for multimodal AI
The patch repairs a defect in the generate process that made some inputs lose their original type and caused problems with multimodal models.
Hugging Face introduces TorchNeuron for AWS's Trainium chips
Hugging Face says running its models on AWS Trainium is easier than ever.
Cline v0.0.48 fixes token bug behind rejected GPT-6 requests
Default models also change for several providers, but the release notes do not say whether existing settings are affected.
Cline CLI v3.0.73 fixes bug that rejected all GPT-6 requests
With no model chosen, the Cline provider uses Claude Sonnet 5.5 again, and default models change for some other providers.
Ollama v0.40.6 introduces q2_0 tensor and file types to GGUF
The release notes do not say what clef is or what VRAM q2_0 requires.
Transformers v5.20.0 breaks code with deprecated cache calls
Code that calls the removed cache methods needs changes, and the release notes do not say how far expert-parallel inference has slowed.
Cursor cloud environments now support up to 300 repositories
An Environments API beta and link sharing for canvases arrived with the update, but no general availability date has been published.
FLUX.3 Image Klein comes to Hugging Face's Diffusers library
Hugging Face announced the addition through the title of a blog post published on 15 Oct 2026.
Gemini CLI v0.64.0 makes file writes and state saving atomic
Chat history storage also changes, but the release notes do not say how much memory and disk efficiency improves.
NVIDIA details NeMo Gym method for tuning chip design agents
NVIDIA says the risk of plausible but wrong outputs spreading through multistage design workflows is a major barrier to production use.
NVIDIA releases full Run:ai GPU platform as open-source code
The move makes production GPU orchestration the open foundation of NVIDIA's DSX AI Factory Platform.
Cursor opens Origin app marketplace to developer tool makers
Origin now has more than 20 integrated apps and will review outside developer tools for listing, but the changelog gives no review time or listing fee.
NVIDIA introduces ReplaySSM for faster hybrid-model decoding
The post examines state storage in models such as Nemotron 3 and the Qwen family, which mix SSM and attention layers.
Devin Desktop 3.12.28 brings per-session MCP server controls
The release also fixes local-session and Remote SSH errors, but the changelog lists no vulnerability numbers for its security fixes.
174 in totalWritten by Jiyul SeoRSS feed




