What’s new at BuildMidas.
New capabilities, useful improvements, and fixes. Follow the progress of AI builders, specialist teams, and game labs, week by week.
19 weeks shown
No updates match your search. Try another term or choose All.
Clearer build progress and a refreshed public site
This week’s work improved verification controls, build feedback and the pages used to explore BuildMidas.
Choose how verification runs
Builds became usable at the build milestone, with controls to disable verification or stop it while keeping the completed version.
Follow progress and credit usage
Version cards gained milestone thumbnails and consumed-credit information. Conversations also gained an updating progress message for each build.
Explore the refreshed landing pages
The homepage, builder and team pages were redesigned, followed by the game labs, blog and detail-page presentation. Existing articles and showcase projects were retained.
More reliable data and creative builds
Fixes addressed workspace ownership failures in data projects and validation problems that could reject valid language tags in AI-agent and content deliverables.
AI teams for more of the work
The platform expanded its business teams and added more kinds of builder output.
Manage social and advertising work
Social and Advertising gained broader account connections, reporting and agent workflows. Supported actions are governed by account settings, review steps and execution limits.
Bring messaging, deals and jobs into the workflow
New team workflows connected inbox work, proposal preparation and job-search tasks with their own interfaces and approval controls.
Work on sales and ecommerce
The Sales team added company research and outreach workflows. Ecommerce connected demand research, storefront building, catalog work and order operations.
Build data projects, AI agents and creative work
New builder workflows cover data and machine-learning projects, deployable AI agents, design work and written content, with verification suited to each kind of output.
Create images and video as deliverables
The builder gained dedicated image and video output workflows, with production and verification handled as part of the build.
Share market evidence across teams
A common market layer made public-source research available to more teams, with shared insights instead of separate fetching in each workflow.
Watch the build, and navigate a clearer workspace
Live build previews and a redesigned dashboard shell made ongoing work easier to follow.
See visual progress while a build runs
Live build watching added preview frames and visual progress tiles during generation and verification, with a dedicated live-view panel.
A clearer dashboard structure
The dashboard moved to a section-based sidebar with subpages presented as tabs, followed by refinements to headings, navigation and spacing.
Inspect more kinds of build files
Code mode improved its handling of engine projects and binary previews, making large game-build output easier to browse.
A more focused enterprise offering
The enterprise page was repositioned around bespoke AI engineering and project requirements.
More engines, more places to play
Engine choice expanded across web, mobile and desktop games.
Godot and Defold become broadly available
Godot and Defold access expanded to all users, while Unity retained its existing access limits.
Choose the engine for the game
2D and 3D build guidance stopped assuming a single default renderer. Mobile and desktop game requests gained the same engine-selection approach.
Engine-backed mobile and desktop games
Godot and Defold game output gained app-shell support. Unity desktop work added native players, followed by an Android artifact path.
Explore Defold in 3D
Defold’s builder and lab scope expanded beyond 2D to include 3D work.
A home for the game labs
A shared Game Lab hub made the engine and asset experiments easier to explore.
Explore labs from one hub
The Game Lab overview and navigation brought Unity Lab, three.js Lab and Blender Studio together in one place.
Godot and Defold get their own labs
Dedicated Godot and Defold galleries introduced each engine’s work. Their initial builder integrations were still gated experiments at this stage.
Establish the playfield before adding art
Game-building guidance added explicit blockout and game-feel stages before final artwork, helping separate layout and interaction decisions from asset production.
Research the opportunity before building
Discovery connected market research, validation and promotion with the builder.
Turn research into Opportunity Briefs
Discovery introduced an AI research workflow that turns an idea or direction into Opportunity Briefs for deciding what to build.
Validate demand and track ventures
Demand validation added a landing-page test and a scorecard, while a venture portfolio connected research, validation and build progress.
Draft promotion for the right platform
Social marketing agents added platform-specific drafts. A Reddit connection brought account-based publishing into the promotion workflow.
See store performance after shipping
Store Analytics introduced charts and tables for shipped-item metrics, alongside a more complete Published and Shipped hub.
From a finished build to a store listing
Store submission, listing material and published-site measurement became more connected.
Ship from the dashboard
The Ship to store workflow added Google Play, Chrome Web Store and Firefox Add-ons submission paths, using the connected store accounts.
Prepare the listing, not just the package
Shipping gained store-assistance kits, listing graphics and copy, and clearer finish-on-the-store steps for work that remains in the store console.
Search and traffic setup for published sites
Published builds gained SEO-file generation and analytics support on their own subdomains, with sitemap submission tied to the published site.
Safer repeat submissions
Re-shipping gained a confirmation guard, and Chrome extension uploads gained recovery for version numbers that had already been published.
Real 3D worlds and an experimental Unity path
The game builder gained a stronger foundation for 3D scenes and authored assets.
Build true 3D games
The builder added a three.js and physics-based game path, with guidance for creating actual 3D worlds and interactions.
Explore Unity game building
An experimental, access-limited Unity WebGL build path introduced a platform-owned build harness and a full-window game player.
Bring Blender-authored assets into games
The asset pipeline gained support for moving Blender output into three.js and Unity projects, connecting authored models with game builds.
Read the game’s actual self-test results
Verification was corrected to collect game self-test results that had previously been missed by the browser checks.
An assistant that can help you take action
Confirmed actions, deeper testing and reusable showcase builds made the workflow more practical.
Act from the AI assistant
The assistant gained handoff cards and confirmed actions for building, publishing, deployment and SEO work, plus checklists for multi-step plans.
Test how a build looks, sounds and plays
New verification work covered visual quality, game self-tests, audio analysis and independent play probes, broadening checks beyond whether a page simply loads.
Save a showcase build as a template
A published showcase build can be saved into your template library as a starting point for another project.
Check more of an application
Application self-tests and multi-page visual audits extended verification to navigation and form flows.
A broader platform, with a stronger visual identity
Public signup, billing and design direction brought more of the product together.
Accounts, plans and credits
Public signup, billing, referrals and the Plans & Credits interface were added around a credit wallet, bringing account access and usage management into the product.
A design director for each build
A design-director pass introduced visual briefs that adapt to websites, apps, extensions and games, with more attention to layout, palette and visual variety.
A dashboard for the whole build journey
The dashboard expanded beyond SEO to show building, shipping and credit activity. BuildMidas also introduced its cube logo and matching favicon.
Native previews and install actions in the showcase
Showcase work restored native-app previews and added an install modal for downloadable builds.
More verification, and games you can publish
Build checks expanded while multiplayer games gained a publishing workflow.
Checks for code, links and accessibility
Verification expanded with code review, security checks, requirement conformance, links, SEO and accessibility checks. Findings became part of the build review process.
Publish multiplayer games
The Games library gained a publish action and a public play route, with runtime handling designed to keep published games available between sessions.
A smoother publishing experience on phones
Published-build actions became a horizontally scrollable row, and publishing dialogs were adjusted to avoid clipping the subdomain suffix on smaller screens.
Prepare browser extensions for stores
Browser-extension publishing gained a store submission pipeline alongside the extension-store account settings.
Preview the product in its own shape
Phone, desktop and extension previews made builds easier to inspect before downloading.
Device-aware previews
Build previews gained browser, phone, tablet, desktop and extension frames, with appropriate viewport controls in templates and build versions.
Scan to install an Android build
Android downloads gained a QR-based install flow with guidance for opening the package on a phone.
Enhance an idea into a brief
The Enhance action added an AI-assisted step for turning a rough idea into a more structured build brief.
Inspect code alongside the preview
The code view gained line numbers and source-anchored navigation, while fullscreen previews were centered more consistently.
More ways to build and organize AI teams
Shared skills, scheduled work and new application formats expanded the platform.
Shared knowledge, skills and schedules
The Agents Hub introduced shared skill attachments, knowledge browsing, per-task schedules and in-app notifications, bringing more of the agent workflow into one place.
Publish dynamic applications
Dynamic applications gained a dedicated runtime and publishing workflow, extending the builder beyond static output.
Build and download Android apps
Native mobile work added Expo previews, Android compilation and downloadable build artifacts. The builder also added browser-extension and Electron desktop build paths.
Organize AI work around tasks
Task views and localized reporting made ongoing agent work easier to follow.
A task-based view of agent runs
A new task-configuration interface introduced reusable work definitions, run details and a docked Coordinator conversation. This was introduced behind the task-configuration feature switch.
Reports in your dashboard language
Agent reporting gained a localization pass, and the selected dashboard language became a saved account preference.
Run numbers that belong to your workspace
Run numbering became specific to each tenant, making a workspace’s first and subsequent runs easier to recognize.
More consistent analytics and clearer deployments
Reporting and deployment feedback became easier to trust and act on.
Keep analytics and agent history consistent
Analytics synchronization moved to its newer data path, while fixes aligned recent actions and run history with their latest status.
Deployment questions in your language
The deployment agent can present its host-choice question in the selected interface language. Connection failures also gained clearer user-facing explanations.
Handle more server layouts
Deployment tooling gained support for different nginx directory layouts and clearer handling of unsupported server setups.
A clearer record of SEO actions
A focused update to reviewing and applying agent recommendations.
Review recent actions before applying them
The SEO workflow gained a Recent actions tab and a manual-apply control, making recommendations and their application easier to distinguish.
Better briefs and builds that check themselves
Structured questions, browser checks and an automatic repair loop improved the path from an idea to a working site.
Clarify the idea before building
The builder gained interactive question cards, including multiple-choice answers and custom responses, so a brief can be made concrete before implementation.
Browser verification and automatic fixes
Completed builds gained browser checks, a visible verification report and an automatic correction loop that can repair problems found in the result.
Build framework-based websites
Astro builds gained a dedicated build path, with previews, publishing and deployment using the generated output rather than the source project.
See what SEO changes achieved
Recent agent reports and impact tracking made it easier to inspect applied changes and their subsequent search-performance signals.
A guided path to your own server
Deployment became a conversation with checks, questions and recovery.
One deployment flow from upload to verification
The deployment agent brought server preparation, uploading and live-URL verification into one workflow, with progress reported as the deployment runs.
Answer deployment questions in place
When a deployment needs a decision, the interface presents the question and resumes from the answer. Domain takeover work also gained an explicit revert action.
More dependable deployment recovery
Deployments interrupted by a server restart are identified on startup, and restoring an earlier version checks that the target version can actually be served.
The first recorded chapter
Publishing, deployment targets and more control over SEO work.
Publish and share a build
Publishing to a BuildMidas subdomain and a public showcase became part of the builder workflow, alongside a single place to find generated builds.
Manage deployment targets
Save a deployment target, attach domains, and choose a separate remote path for each site. Version history added a way to revisit earlier deployments.
Run SEO agents for a specific domain
Domain actions gained individual agent runs and domain-specific settings, giving users a more direct way to start and inspect SEO work.