Changelog
Stay up to date with new features, improvements, and bug fixes shipped in Vybe.
Stay up to date with new features, improvements, and bug fixes shipped in Vybe.
June 15, 2026
A big week. You can now see exactly what your crons are doing, pick models with full transparency on cost and capability, and navigate the platform from a redesigned sidebar. Plus: Fable 5, Gemini Flash High, and the ability for agents to own their inbound email and SMS.
Your crons now have memory. Trigger any task on demand, browse every past run with its status and duration, and drill into a specific run to see exactly what the agent did and what tools it called. Per-cron model overrides included.
The model selector got a full overhaul: detail cards show intelligence scores, credit costs, and context window sizes. Models are grouped and sorted so you can actually make an informed choice rather than guessing from a name.
When a model is thinking, you can see it. Reasoning steps and Gemini thought summaries now surface directly in the chat, so you know what is driving the response — not just the answer.
Three new models added this week. Claude Fable 5 is now in the standard stack, Fable 5 XHigh is in beta with 128K output token support, and Gemini 3.5 Flash High lands as its own distinct option.
You can configure an agent to handle messages that arrive at its inbox — no cron polling required. Email and SMS inbound handling is now a first-class setup option.
Shared Postgres databases are now available at the org level. Multiple agents and apps can read and write the same data without duplicating state across each of them.
The sidebar has been redesigned with a floating nav, dedicated submenus for AI coworkers and apps, and a refreshed header. Cleaner to scan, faster to get where you are going.
Dropped connections no longer mean lost responses. The platform now resumes or reconciles persisted messages when a stream cuts out, so an interrupted run picks back up cleanly.
Ready to try these new features?
June 8, 2026
This week adds the start of a real permissions model for agents, gets agents replying to emails by default, and puts cron management where it should have been all along. Plus a lot of UI fixes the team had been stacking up.
Owners and admins can change member roles directly from org settings.
Admins can now resend or cancel pending invitations.
When an agent gets an email, it replies to the owner automatically. No config needed.
Create, update, and delete crons from the background tools panel. No more needing a separate workflow to manage schedules.
You can rename note titles directly in the memory panel.
Agents can now be assigned admin roles, giving them the ability to act on behalf of org admins. Useful if you have an agent managing other agents.
The askUserQuestion tool is no longer exposed to agents running in Slack channels, where it was creating unnecessary prompts.
Loading states, thread messages, markdown rendering, onboarding menus, usage dashboard, and card layouts all got fixes this week.
SQL errors from runSql and agent databases now show the actual Postgres or Neon error, not a generic fallback.
Ready to try these new features?
June 1, 2026
Agents can now see and control their own event triggers. That's the big one this week. Alongside that: Claude Opus 4.8 is live, credits are easier to add, and a run of fixes cleans up Slack reliability, checkout billing, and email reading.
Agents now know what event triggers they have and can create, enable, disable, and delete them on their own. Tell your agent what events to watch for and it wires itself up.
Added at medium and xhigh intelligence levels. A solid option for agents doing heavy reasoning or complex multi-step tasks.
A new modal with a slider makes adding credits fast. No context switching, no separate flow.
When publishing fails, the error comes back into the chat. The agent can diagnose and retry without you digging through logs.
The Slack panel setup content is cleaner. First-time connection feels like onboarding instead of reading documentation.
Connecting a custom sending domain has fewer requirements now. Less friction when setting up agent email.
The agent integration settings dialog is reorganized and easier to scan when managing multiple integrations.
Two issues patched: agents now favor DMs correctly when a channel message is denied, and Slack actions only mark as sent after the tool call actually succeeds.
Credit allocation and trial day defaults were wrong on certain paid plans. Both are fixed.
Creating notes from inside an agent session was failing in some cases. Works cleanly now.
Ready to try these new features?
May 25, 2026
The usage dashboard is live. Gemini 2.5 Flash is available as a model. Slack connection flows work better. And a few bugs that were quietly breaking things got fixed.
You can now see what your agent has been doing. The dashboard defaults to the last 7 days and breaks down activity by tool calls and conversations.
Gemini 2.5 Flash is now available as a model option. It is faster and cheaper than most alternatives, which makes it a good fit for agents running high-volume or repetitive tasks.
Agents can now trigger the external database connection flow directly from their tool panel. You no longer have to navigate away to set one up.
Two agents can no longer share the same inbox address. The system enforces uniqueness at setup, so you catch the conflict before it causes problems.
Tool categories in the capabilities panel are labeled now. Easier to find what you are looking for without having to open each section.
After connecting Slack, agents now open the relevant subthread automatically. There is also a new tool that lets agents surface the Slack install button directly in conversation.
Creating a cron job that already exists used to fail silently. Now you get an explicit message so you know what happened.
Fixed a visual glitch where the note link popover would flash when switching tabs. Small fix, but noticeable.
Ready to try these new features?
May 19, 2026
A focused week on first impressions and visibility. Apps now get AI-generated icons on creation, the home launcher got rebuilt from scratch, and a usage dashboard makes it clear where your credits are going. Plus domain-level signup blocking and deploy tools for agents.
Apps get a custom icon when created. You can also regenerate it anytime from the agent.
The setup panel is gone. Replaced with a Spotlight-style home launcher — faster to navigate, less in the way.
You can now see where your credits are going, broken down by agent and run.
Two new tools let agents kick off app deployments and check their status without leaving the conversation.
Agents can handle any file type attached in a conversation, not just a handful of supported formats.
Settings sidebar now has sub-items so you spend less time hunting for the right page.
More reliable enforcement of which senders can trigger your agent by email.
Cleaner layout, tighter grid, and the creator and last-edited metadata is gone.
New agents get an avatar that matches the gender of their name. Small thing, but it matters.
Active meetings now show a pulse in the agent interface so you know a call is being recorded.
Ready to try these new features?
May 10, 2026
A dense week of UI and infrastructure work. The Mac dock got a full redesign, brand themes landed, the avatar gallery expanded, and the foundation for per-agent databases shipped. Plus BCC support, wiki-style note links, and WhatsApp access controls.
Every agent now gets its own dedicated Postgres database. Persistent, isolated, and queryable directly from agent code via the queryAgentDatabase tool.
The agent dock got a full redesign — cleaner layout, better hierarchy, and faster to navigate between agents.
Agents can now reflect custom brand colors and visual identity throughout their interface.
More avatar options on agent creation, with better gender and style coverage.
Notes now support [[slug|Display Text]] links to other notes and skills, building a connected knowledge graph across your agent.
Agents can now BCC recipients when sending emails.
Tighter controls over which contacts can trigger an agent via WhatsApp.
Each scheduled task can now run on a different model. Useful for cost and quality tradeoffs across jobs.
Ready to try these new features?
May 5, 2026
A focused week on trust and reliability. Agents now have inbound/outbound email contact controls, the agent gallery got a full redesign, and sandbox execution saw significant stability improvements.
Agents now have inbound gate and outbound tool enforcement — control exactly who agents can email and who can reach them.
Manage which apps are linked to an agent directly from the Access page, with access-level gating and visibility controls.
Added a proxy execution layer for custom APIs with CLI, agent tools, and app helper support.
Agents now receive the currently viewed app on every message, and can redirect the user preview panel with the new showApp tool.
Agents can now prompt users with an in-product satisfaction survey to collect structured feedback.
When web search providers fail, agents now receive structured fallback guidance instead of a dead end.
Slack responder is now blocked when agent access is disabled. Mention-only response mode is available.
Integration and custom API secrets are now invalidated immediately on soft-delete.
Ready to try these new features?
April 28, 2026
A feature-dense week. The MCP server catalog launched, GPT-5.5 is now available, notes gained full version history with restore, and scheduled task guardrails were introduced to keep crons accountable.
Browse and connect MCP servers from a built-in catalog, now consolidated into the integrations page.
GPT-5.5 is now in the model picker. Older GPT models have been hidden to reduce clutter.
Notes and skills now store up to 50 versions. Roll back to any previous version from the history panel.
Cron jobs now have a contract + self-report + verifier system to ensure they complete expected side effects. Failures surface in the UI.
Cron descriptions can reference skill slugs — skill content expands at runtime so updates flow through automatically.
Slack read tools now return the channel name or DM partner handle alongside messages.
Pressing Escape in nested modals now closes only the innermost one.
Fixed a bug where prices checkout subscriptions were incorrectly auto-canceled.
Ready to try these new features?
April 21, 2026
A landmark week for the platform. Agents can now send platform feedback, MCP OAuth is live, the gallery got a fresh redesign, and a new pricing page launched. Sandbox stability saw major improvements across the board.
Fresh gallery look with a personalize modal and improved sidebar navigation for discovering and launching agents.
A new dedicated pricing page is now live with clean query-parameter routing for deep-linking.
MCP servers now support OAuth-based authentication, with a connect confirmation dialog shown after creation.
Claude Opus 4.7 is now available and set as the new default where Opus 4.6 was previously used.
Agents now have a platform feedback tool to flag bugs, capability gaps, and tool quirks directly to Vybe engineers.
MCP servers are now subject to the same access control and permission checks as other integrations.
App sandboxes inactive for 29 days are automatically expired to free up resources.
Multiple fixes to sandbox process timeouts, error handling, and H2 client timeouts for more reliable code execution.
Ready to try these new features?
April 12, 2026
Notes now support full folder organization. You can lock notes to prevent accidental edits. Agent templates are live. The cron limit jumped from 10 to 50. We also added a stop button for streaming and moved scheduled tasks to their own sidebar page.
Notes now support full folder hierarchy. Create subfolders, move notes between folders, and keep agent memory organized at scale.
Notes can now be locked to prevent accidental edits. Skills are always locked by default.
You can now create agents from templates. Speeds up setup for common agent types.
The scheduled task limit jumped from 10 to 50 per agent. If you were hitting the ceiling, you are not anymore.
You can now stop an agent mid-stream. No more waiting for a runaway response to finish.
Ready to try these new features?
April 5, 2026
The agent sandbox is fully operational with lifecycle management and the executeCommand tool. WhatsApp is live via Kapso. SMS auto-reply works. Agents can now contact other agents directly. Platform skills shipped too.
The agent sandbox now has full lifecycle management. The executeCommand tool is live, giving agents the ability to run shell commands in an isolated environment.
Agents can now auto-reply to inbound SMS messages.
Agents can now contact other org-level agents directly using the contactAgent tool.
Platform-level skills are live. These are org-shared instruction sets that persist across agent versions.
WhatsApp is now available as an agent channel via Kapso API. Agents can send and receive WhatsApp messages.
Ready to try these new features?
March 29, 2026
Each agent now supports its own custom instructions. Slack contact restrictions let you control which agents can DM users. Pipedream action validation is live: inputs are checked against configurableProps at runtime before any call goes out.
Each agent now has its own custom instructions field. Useful for giving specific agents context or constraints that should always be in scope.
Admins can now restrict which agents are allowed to DM users in Slack. Prevents unwanted outbound contact from automation agents.
Notes now have a graph view that shows connections between notes visually.
Action inputs are now validated against configurableProps at runtime before any external call is made. Catches bad inputs early.
You can now favorite agents. Favorited agents show as cards, same visual treatment as apps.
Ready to try these new features?
March 22, 2026
Agents now have granular per-integration account permissions. Email auto-reply is live via webhook. We shipped a dedicated communications page that puts email, calendar, meetings, Slack, and phone in one place.
You can now grant agents access to specific integration accounts rather than all-or-nothing. Useful when one org has multiple Slack or HubSpot connections.
Agents can now listen for inbound emails via webhook and auto-reply. Set it up once, it runs on every new email.
Agents can now participate in Slack group DMs, not just 1:1 conversations.
You can now upload a transcript manually to attach to a meeting record.
Email, calendar, meetings, Slack, and phone are now consolidated into a single agent communications page.
Ready to try these new features?
March 15, 2026
Skills are live. Agents can store reusable instruction sets and load them on demand. We also shipped meeting search, a context window usage indicator, and a dedicated memory page. Access controls got a full phase-3 upgrade with granular per-integration permissions.
Agents can now create, store, and load reusable instruction sets called skills. The skill is loaded fresh each time, so updates propagate automatically.
Agents can now search past meeting transcripts by title, speaker, or content.
A usage indicator now appears in chat so you can see how much context the agent is consuming per message.
Agent notes now live on their own page with direct URL navigation. Easier to find, edit, and share.
Scheduled task descriptions render markdown, including inline note links.
Ready to try these new features?
March 8, 2026
Granola is now a built-in integration. We added a dedicated code explorer sub-agent for read-only codebase research. Under the hood, the agent runtime got unified: chat, heartbeat, and SMS all run through the same core now.
A dedicated read-only codebase research agent. Agents can delegate investigation tasks without risking accidental writes.
Agents can now summarize threads and branches. Useful for long-running conversations where context needs to be compressed.
Chat, heartbeat, and SMS all run through the same core execution engine now. Behavior is more consistent and easier to test.
You can now give integration accounts human-readable names instead of relying on default labels.
Granola is now available as a native integration. Meeting notes flow directly into agent context.
Ready to try these new features?
March 1, 2026
We ripped out Nango entirely and rebuilt integrations on our own stack. Agents can now receive images in chat. Context management got a serious upgrade: compaction, reset on new messages, and a heartbeat step limit that jumped from 10 to 30.
Agents can now receive images sent in chat. Works in both Vercel Blob storage and the agent thread UI.
We cut the Nango dependency entirely. Integrations now run on our own stack, which means fewer moving parts and no more Nango-related auth issues.
Context compaction is live, along with proper reset behavior on new top-level messages. Heartbeat step limit also raised from 10 to 30.
The preview navbar now has a fullscreen toggle. Small UX thing, but agents that build apps needed it.
Agents in Slack now stay assertive. The previous behavior where they could loop into clarification questions is gone.
Ready to try these new features?
February 18, 2026
We're excited to share what's new in this week's update. Here are the highlights:
Ready to try these new features?
February 10, 2026
This update adds expanded AI model support, including GPT Codex, along with improved SSO controls. Organizations can define trusted SSO domains and set custom AI instructions that apply across teams. Google Drive integrations now support listing and sharing files directly within workflows.
Organizations can now set custom AI instructions that apply to all users within the organization. This feature allows teams to tailor AI behavior to better fit their unique workflows and communication styles.
We introduced the new frontier model Opus 4.6. What’s new with Opus 4.6: - Stronger in large codebases apps: stays coherent where previous models could get overwhelmed (1m token) - Adjustable reasoning effort including very long - Better management of compaction of context when it goes above 1m - Better planning for long-horizon work
We've modernized the interface with a streamlined sidebar, improved transitions, and updated icons for a cleaner, more intuitive experience.
You can now configure trusted SSO domains, making it easier and more secure for users to sign in through approved identity providers.
SSO enforcement has been added, requiring all users to sign in using Single Sign-On.
Made additional enhancements to the workflow interface for smoother navigation and usability.
When adding a custom API, Vybe now automatically validates your OpenAPI specification. This helps catch errors early and ensures your API integrations work reliably.
Resolved a problem that prevented invitations from being created correctly. Users can now send and receive invitations without errors.
Enhanced the updatefile tool to provide clearer and more actionable feedback during file updates.
Ready to try these new features?
January 28, 2026
We now support SAML and OIDC authentication support. And as usual, we shipped tons of new functions on HubSpot (retrieve multiple deals with our new batch action) or GraphQL queries on Monday.com as well as new Mixpanel actions.
Chat messages now feature a cleaner, more compact collapsed view and an animated icon that shows real-time streaming progress.
You can now log in using SAML or OIDC identity providers, making it easier to connect with your organization’s SSO solutions. This will enhance security and streamlines access for enterprise users.
The Integrations page now features a new Table View for each section, making it easier to view and manage your integrations.
You can now softly delete apps instead of permanently removing them. This feature helps prevent accidental loss by allowing recovery of deleted apps.
Users now receive clear error messages if sending an invitation fails.
The editor and chat interface have been updated with improved background colors, cleaner gradients, and subtle borders for a more polished look.
We've moved our sandboxes and Neon projects to be geographically closer to the core vybe application. Expect faster performance on deployed apps.
The sandbox preview iframe now allows clipboard writing, resolving errors when user apps try to use the Clipboard API.
Enhanced the look and feel of the Editor Tab Bar by refining its hover and selected states for better visual feedback.
Messages from the AI now fully support Markdown formatting.
Resolved a bug that prevented viewers from accessing deployed applications.
Updated the look and feel of feature suggestions.
Ready to try these new features?
January 23, 2026
Our agent now asks for clarification, using interactive options to ensure you get more precise and helpful responses. We also added 30 new custom actions for Notion, HubSpot, Slack, and Google. You can now create Notion pages, update HubSpot meetings, and send emails via Gmail.
Introduced a new feature (currently behind a feature flag) that allows the agent to ask users clarification questions using select and multi-select options, improving interaction flow and ensuring more precise responses.
We’ve improved the OpenAPI integration by refining tool names and upgrading the OpenAPI directive for clearer, more intuitive use. Additionally, a new tool has been added to automatically generate an OpenAPI client for your custom APIs.
The Data panel now features a new Integrations tab along with simplified horizontal tab navigation. This makes it easier to view and manage all your connected services and databases in one place.
Chats now load more quickly by displaying recent messages first and loading older messages as you scroll. This improves performance, especially in conversations with many messages, so you can start chatting sooner.
The chat tool now includes an AI-assisted type checking feature, offering quick type verification without needing a full build. While it's currently slower for larger apps, this update helps catch type errors earlier to streamline development.
Tool outputs that return JSON are now shown in a clearer, easier-to-read format thanks to a new reusable UI component. This update also fixes imports in the Gmail Inbox component for better reliability.
We've improved the quality of AI-generated suggestions to provide more accurate and helpful guidance. This update helps you work more efficiently and make better decisions within the app.
Ready to try these new features?
January 15, 2026
We're excited to introduce a bunch of updates making Vybe Integrations even more functional and user-friendly. From supporting multiple Lemlist endpoints to new Google Gemini image generation, we've enhanced workflow automation and data management across various platforms to help you work smarter and more efficiently.
You can now explore a new Templates page in the app, making it easier to browse and use ready-made templates.
Added persistent AI suggestions that are saved with your project, allowing you to access and revisit suggestions even after reloading or returning to your app for consistent guidance throughout your workflow.
You can now use a 'Test Run' button on workflow functions to quickly perform ad hoc test runs. This makes it easier to validate your workflow setup before deploying changes.
Added an integration icon bar to the chat input, making it easy to @mention connected apps, and improved the mentions dropdown with a "Connect more apps" option to simplify integration management.
The appearance of build error messages has been simplified for a less intrusive experience during builds. This change helps you stay focused in chat without overwhelming error alerts.
Ready to try these new features?
Become an AI native company
AI coworkers and custom apps that run your operations
Vybe, Inc. © 2026