Every assistant inbox now has a handle you control. Instead of the auto-generated address you got when email was first enabled, you can pick something short and memorable and hand it out with confidence. The change applies to both OpenClaw and Hermes assistants.

• Choose your own handle: On the Email tab for any assistant, type the name you want before the @ and save. Your inbox address updates immediately and any future email sent by the assistant goes out from the new address.
• Change it whenever you want: Renamed your assistant, or just want something catchier? Update the handle again and the new address takes effect right away. No redeploy to babysit on OpenClaw, and a seamless redeploy on Hermes.
• Globally unique, guaranteed: Handles are reserved the instant you claim one, so two assistants can never share an inbox. If someone else has already taken the name you want, you'll see a clear message and can try another.
• Friendly validation as you type: The form tells you right away if a handle is too short, uses unsupported characters, or touches one of the reserved mailbox names. No guessing, no trial-and-error saves.
• Defaults unchanged if you don't care: Assistants that already have email enabled keep their existing address — nothing moves until you choose a new handle.
• A cleaner delete confirmation: Deleting an assistant now opens a proper in-app dialog instead of a browser prompt, with a loading state while the tear-down runs and a smooth exit animation on the assistants list.

A random identifier is fine for an internal ID, but it's a terrible thing to hand out to colleagues, customers, or vendors. Now your assistant's inbox can look like it belongs to a person, not a database row.

Hermes is now more than 50% faster. Below is time to first reply — how long between a user's message and the assistant's first word back — tracked over the last few weeks. The baseline held steady for a long time; this release is the cliff on the right.

Two shifts landed in this release. Replies are dramatically faster (above), and assistants now keep their full state across redeploys — every conversation, uploaded file, scheduled job, and approved pairing survives.

• Over 50% faster replies: Simple questions that used to take about 30 seconds now come back in around 7. Complex tasks that took a full minute now finish in roughly 20.
• Automatic snapshots on redeploy: The moment you trigger a redeploy, your assistant's state is captured and saved before the running instance is replaced. Nothing extra to configure, nothing to remember.
• Seamless restore on the new instance: When the replacement instance comes up, it boots with the exact workspace, memory, and channel pairings the previous one had. Conversations pick up right where they left off.
• Existing assistants migrated automatically: Hermes assistants created before this change have their prior data moved over into the new snapshot format, so long-running assistants keep their history without any manual work.
• More forgiving redeploys: A redeploy no longer fails just because your assistant is busy working through a backlog or waiting on an external provider. The platform gives it room to settle instead of restart-looping.
• Faster channel reconnection: After a redeploy, Telegram pairings and other approved channels are already restored, so there's no re-pairing or re-authorization step.

Redeploys used to be the scariest button in the product for long-running Hermes assistants, and waiting on replies was the slowest part of a typical day. With this release, shipping an update is routine instead of a data-loss risk, and your assistant answers while you're still looking at the screen.

When you create a managed assistant, the model picker now pulls in the full OpenRouter catalog instead of a short curated list. That means you can browse and search across 500+ models from dozens of providers, all in the same dropdown. We also updated the defaults so new assistants start on the latest models.

• Full model catalog: The model selector during onboarding now shows every model available through OpenRouter, organized by provider. Popular providers like Anthropic, OpenAI, Google, and DeepSeek appear first, and you can type to search the rest.
• Instant search across all models: Start typing a model name, provider, or ID and the list filters down immediately. No more scrolling through a fixed list hoping the model you want is there.
• GPT-5.4 Mini as the new default: Managed assistants now default to OpenAI's GPT-5.4 Mini, which offers strong performance at a lower price point. You can always change it before confirming.
• GPT-5.4 and GPT-5.4 Mini added: Both models are now available for managed and bring-your-own-key assistants, so you can use them regardless of how your assistant is set up.
• Context length at a glance: Each model in the dropdown shows its context window size, so you can make an informed choice without leaving the setup flow.

This update makes it much easier to find and use the right model for your assistant, whether you want the latest flagship or a cost-effective option for simpler tasks.

Hermes assistants now support inbound email. When someone sends a message to your assistant's email address, it arrives as a structured webhook event that the agent can process, respond to, and take action on. This brings Hermes to parity with OpenClaw on one of the most requested messaging channels.

• Email forwarding for Hermes: Inbound emails are routed to Hermes assistants through the same address-based forwarding that OpenClaw assistants already use, so setup works the same way.
• Structured webhook delivery: Hermes receives emails as typed webhook events with the sender, subject, body, and message threading information preserved, so the agent has full context to work with.
• Persistent storage for Hermes containers: Hermes assistants can now use durable storage that survives container restarts, so files, state, and workspace data stick around between sessions.
• More reliable terminal startup: The health sidecar now lazy-loads the terminal emulator, avoiding startup failures when the native module isn't immediately available inside the container.

Email is one of the simplest ways to interact with an assistant without installing anything. With this update, both OpenClaw and Hermes assistants can be reached the same way.

Hermes assistants now come with a built-in terminal in the dashboard. That means you can open a live shell directly inside the assistant's container, inspect files, run commands, and debug issues in place instead of guessing from the outside.

• A dedicated Terminal tab: Active Hermes assistants now have their own Terminal tab on the assistant details page, so shell access is part of the normal management flow instead of a separate workaround.
• Live shell access inside the container: You can connect to the running Hermes environment, execute commands, inspect the filesystem, and verify what is actually happening on the deployed assistant.
• Faster debugging when something feels off: If an assistant is behaving unexpectedly, you now have a direct way to check logs, files, processes, and runtime state without waiting for a redeploy or trying to infer everything from the UI.
• Secure, short-lived terminal sessions: Terminal access is gated behind signed session tickets and tied to the active assistant session, so it is designed for controlled troubleshooting rather than permanent open access.
• A more dependable Hermes runtime underneath: We also tightened the runtime path behind the scenes so terminal sessions connect more reliably and deploy status updates are less likely to fail during transient errors.

This is the most direct way we have added yet to work with a hosted assistant like a real running system, not just a dashboard object. For Hermes users especially, it makes day-to-day debugging and inspection much more practical.

getclaw now supports Hermes Agent as a hosted framework. That means you can create a Hermes assistant directly from the product, deploy it through the same managed flow, and operate it from the same dashboard you already use for OpenClaw assistants.

• A second hosted framework: You can now choose Hermes Agent instead of OpenClaw when creating a new assistant, directly inside getclaw.
• Managed through the same dashboard: Hermes assistants live inside the same deployment, management, and post-launch workflow as the rest of the platform, so there is no separate hosting path to learn.
• Telegram support for Hermes deployments: Hermes assistants can now run on Telegram in the hosted environment, so messaging works as part of the normal getclaw channel setup flow.
• Approval controls for Telegram access: When a restricted user messages a Hermes assistant, their Telegram user ID shows up in the dashboard so you can approve them without leaving getclaw.
• A more usable hosted Hermes runtime: Hermes deployments now boot cleanly in the managed environment, which makes hosted Hermes much closer to a real day-to-day option instead of an experimental add-on.

This is a framework expansion, not just a new toggle. getclaw now supports two different open-source agent runtimes, which gives users a clearer choice between OpenClaw's operations-first model and Hermes's more personal, persistent-agent model.

Until now, each conversation with your assistant started from a blank slate. The assistant could be brilliant in the moment, but the next time you talked to it, all of that context was gone. Lossless-claw changes that. It gives every managed assistant a persistent, structured memory that grows over time — so your assistant gets more useful the more you use it.

• Full conversation memory: Lossless-claw captures the complete context of every conversation — not just summaries or keywords, but the actual substance of what was discussed, decided, and learned.
• Automatic context recall: When your assistant starts a new conversation, it pulls in relevant context from past interactions. It knows what you've worked on, what you've asked before, and what preferences you've expressed — without you having to repeat yourself.
• Gets smarter over time: The more you use your assistant, the better it understands your projects, your terminology, and how you like things done. This is the difference between a tool that resets every session and one that actually learns.
• On by default, no setup: Every assistant now ships with lossless-claw pre-installed and active. There is nothing to configure. If you need to disable it for a specific assistant, set the DISABLE_LOSSLESS_CLAW environment variable.
• Baked into the image: The plugin is pre-installed in the container image at build time, so there is no download delay on cold starts and no risk of install failures at runtime.

This is a step change in what managed assistants can do. An assistant that remembers is fundamentally more useful than one that forgets — for ongoing projects, repeated workflows, and anything where continuity matters.

The infrastructure underneath managed assistants is now faster to start, easier to debug, and lighter on resources. This release removes unused dependencies from the container image, parallelizes data restores, and finally gives you visibility into what actually went wrong when a deploy fails.

• A smaller, faster container image: Playwright and pnpm have been removed from the Docker image. The result is a leaner image and faster startup.
• Parallel R2 restores: Workspace and skills data now restore from R2 concurrently instead of sequentially, cutting restore time on restarts.
• Delete backup data from the dashboard: You can now delete individual files or entire directories from your assistant's R2 backup storage directly in the dashboard, with streamed progress for large deletions.
• Safer Telegram defaults: If a Telegram allowlist is configured with no approved users, the runtime now falls back to pairing mode automatically instead of rejecting all messages.

Assistants start faster, and when something breaks during startup, you can actually see why.

Managing channels is now much cleaner, and the runtime underneath managed assistants just took a meaningful step forward too. This release adds a dedicated Messaging tab for channel setup and rolls hosted assistants onto OpenClaw 2026.3.13, which pulls in the biggest recent capability upgrades from upstream.

• A dedicated Messaging tab: Telegram, Slack, Discord, and Email now live in one place on the assistant details page, so channel setup no longer feels scattered across separate screens and controls.
• Easier channel management: You can enable or disable channels, rotate bot tokens, see whether a channel is configured, copy an assistant's email address, and jump straight into email management without leaving the dashboard.
• The big OpenClaw upgrade is now in managed deployments: Moving to OpenClaw 2026.3.13 brings the biggest recent upstream improvements into the hosted runtime, including browser automation, assistant email support, and unlimited built-in search.
• A safer and more predictable runtime: Deployments now validate the OpenClaw config before startup, set Control UI origins more explicitly for hosted environments, and pin Telegram streaming to a safer default so upgrades are less likely to break in surprising ways.

This is partly a dashboard release and partly an infrastructure one, but both sides matter. Channel operations are easier day to day, and managed assistants are now much closer to the current OpenClaw experience instead of lagging behind it.

Search is now available to every OpenClaw user, and it ships with unlimited usage from day one. No plan upgrade, extra toggle, or special setup required. If you already have assistants deployed, just redeploy each instance once to enable the new feature.

• Search is now built in: OpenClaw assistants can now use search as a native capability, opening up research and retrieval workflows that were not available before.
• Unlimited usage included: Every user gets the feature with no search cap, so assistants can keep using it without running into usage limits.
• One redeploy to enable it: Existing instances need a redeploy to pick up the new runtime behavior. Once redeployed, the feature is active automatically.
• Applies to your current assistants: This is not limited to newly created assistants. Your existing instances can use search too as soon as they are redeployed.

This is a meaningful capability release, not just a limit change. Search is now part of what every assistant can do, and if you already have one live, a redeploy is all it takes to turn it on.

OpenClaw assistants now have their own email address. They can check their inbox, send messages, and reply to threads — all as part of a workflow.

• Dedicated email address: Each assistant gets its own email address, provisioned automatically when the skill is enabled. No external setup required.
• Send, reply, and check inbox: The assistant can compose new emails, reply within existing threads, and read incoming messages on its own.
• Generous free tier: Every user gets 100 free emails per month. Beyond that, additional emails cost fractions of a cent from your credit balance.
• Threaded conversations: Replies are linked to the original message, so recipients see a clean conversation thread instead of disconnected messages.

Email turns your assistant from something you talk to into something that can talk to others on your behalf — following up, reporting, and communicating without you in the loop.

OpenClaw can now use a real browser as part of its workflow. That means your assistant is no longer limited to APIs and text-only tools. It can move through the web the way a person would: opening pages, clicking buttons, reading content, and acting on what it finds.

• Navigate real websites: OpenClaw can visit live pages, wait for interfaces to load, and work through multi-step browser flows instead of stopping at static HTML.
• Interact with the page: Clicking, form filling, scrolling, and page-to-page navigation are now part of the assistant's toolkit, which unlocks workflows that were previously impossible with API-only access.
• Read what users actually see: The assistant can inspect rendered pages, not just backend responses, which is a big step forward for research, operations, QA, and support-style tasks.
• Unlock broader automation: This creates room for assistants that verify dashboards, collect information from tools without public APIs, reproduce UI issues, and complete browser-based tasks end to end.

This is one of the biggest capability jumps in the product so far. Giving OpenClaw a browser turns it from an assistant that can call tools into one that can operate directly on the web.

This release is about polish where it matters: assistants start more smoothly, the first-run experience feels more alive, and Telegram access is less likely to surprise you later.

• More dependable launches: Deployments now recover more gracefully during slower startup moments, so assistants are less likely to get stuck while going live.
• Telegram access that stays put: If you maintain a private Telegram allowlist, those approved users now stay approved more reliably across edits and redeploys.
• Clearer save and redeploy feedback: When you update files from the dashboard, you now get much better progress messaging so it's obvious what was saved and what's still rolling out.
• A more premium first impression: Persona previews load faster, hover demos feel smoother, and the homepage greeting feels more intentional instead of generic.

Small release on paper, but it removes a lot of friction in the moments users actually notice: launch, preview, and trust.

Creating an assistant no longer means filling out forms. Just describe what you want in a conversation, and the wizard configures everything for you.

• Chat-based creation: Describe your assistant in plain language — name, personality, model, API key, and channels are all set through a guided conversation.
• 6 pre-built personas: Get started instantly with LinkedIn Influencer, News Reporter, Sales Rep, Customer Support, Software Developer, or Personal Coach — each with a ready-to-go personality and skills.
• Redesigned homepage: A new homepage showcases what's possible with persona demos and video walkthroughs.
• Fresh visual identity: Updated logo, favicons, and branding across the product.

Two improvements to help you understand and manage running assistants faster.

• Live server logs: A new logs panel on the assistant detail page streams your worker's output in real time, with color-coded log levels and auto-scrolling.
• Playback controls: Start, stop, and clear the log stream without leaving the dashboard.
• Faster config changes: Updating your model or API key now takes effect in about 30 seconds, down from a full redeploy that could take several minutes.
• Visual update status: The dashboard shows an "Updating configuration..." indicator while changes are being applied.

Your assistant is no longer limited to a single messaging platform. You can now deploy to Telegram, Slack, Discord, or any combination of the three.

• Slack support: Connect your assistant to a Slack workspace using a Bot Token and App Token. A built-in setup guide walks you through creating the Slack app.
• Discord support: Connect your assistant to a Discord server with a single bot token. The setup guide covers bot creation, permissions, and server invitation.
• Multi-channel deployments: Enable multiple channels on a single assistant. The detail page shows all active channels at a glance.
• Step-by-step setup guides: Each channel includes an expandable guide with the exact steps, scopes, and permissions needed.

Until now, editing assistant files meant working through openclaw UI or redeploying from scratch. The new file editor gives you direct access to your assistant's files right from the dashboard.

• Browse your assistant's files: Navigate directories, open files, and explore your assistant's configuration and workspace from a familiar file explorer.
• Edit with syntax highlighting: A full-featured code editor with automatic language detection for JSON, TypeScript, JavaScript, YAML, Markdown, shell scripts, and more.
• Save & redeploy in one click: Edit a file and redeploy your assistant in a single action, no need to leave the dashboard.
• Safe by default: Unsaved changes are flagged before closing, and saves are validated to prevent conflicts.

This update improves how managed assistants preserve state, so restarts are less disruptive and recovery is more dependable.

• Better data safety: Backups now run more consistently, including periodic full backups to reduce the chance of drift over time.
• Smoother restarts: Assistants now recover newer saved state more reliably when containers restart.
• More complete recovery: Restores now cover assistant state, workspace files, and installed skills.
• Clear backup health: Admin status now shows last successful backup, last backup attempt, and recent backup errors when present.
• Customer value: Less manual recovery work, fewer surprises after restarts, and more confidence that assistant progress is preserved.

We made this change because cloud browser automation setups were too fragile and required too much manual configuration. The goal of this release is to give you safer defaults, fewer deployment steps, and more reliable assistant behavior in isolated environments.

• Why this matters: Cloud environments can fail in ways that are hard to debug when browser connectivity is not configured securely and consistently.
• What you now get: Deployments now apply secure connectivity defaults automatically, so assistants come up with fewer manual steps.
• Reliability improvement: Isolated sandboxes now default to the Playwright CLI browser path, which avoids common extension-relay failures.
• Stability improvement: Gateway configuration is normalized for current OpenClaw runtimes, reducing startup and connectivity issues.
• Customer value: Faster time-to-working assistant, less deployment troubleshooting, and safer browser automation defaults out of the box.

We are introducing first-class support for fully customizing your openclaw.json, alongside making assistant configuration easier to understand and more dependable.

• Customize openclaw.json: You now have direct control to customize your openclaw.json configuration just how you want it.
• Clearer config status: You can now quickly see which settings are currently applied versus still pending.
• Easier troubleshooting: Side-by-side config comparisons make it faster to spot what changed.
• More reliable updates: Config saves are now applied more consistently across assistant redeploys.
• Cleaner comparisons: Runtime-managed values are filtered out so you can focus on user-controlled changes.
• Supported OpenClaw version: 2026.2.19.

This release focused on product clarity and assistant configuration flow.

• Model updates are easier: Change your assistant model directly from the assistant details page.
• Cleaner logged-in experience: Returning users are now guided to their assistants right away.
• Improved onboarding content: New comparison content helps users choose the right path faster.

This update focused on clarity and usability.

• Smoother account flows: Improved key flows so setup and activation feel more reliable.
• Better docs navigation: Updated structure and links so it is easier to find what you need.
• Clearer guides: Expanded and refined help content for common usage patterns.

This release made external integrations more complete while improving day-to-day stability.

• More complete integrations: Added broader support for assistant management tasks.
• Easier setup: Updated guides and behavior for a smoother integration experience.
• Stability improvements: Fixed issues that could affect account, billing, and assistant operations.

This patch focused on making managed usage more dependable.

• More accurate credits: Resolved edge cases around balance updates and top-ups.
• More stable assistant performance: Fixed issues that could interrupt managed assistant responses.

A full pass on the landing experience improved readability and conversion-oriented messaging.

• New page structure: Improved layout and section flow to make the value proposition clearer.
• Visual and copy polish: Updated visuals and messaging to improve readability and conversion.

This was a major milestone for managed assistant billing and reliability.

• Managed credits are live: Users can now run assistants with platform-managed credits.
• Improved account isolation: Credits are now separated per user for better consistency.
• Updated pricing structure: Managed assistant plans were clarified and separated from other offerings.
• Deployment reliability improvements: Managed assistant deployment flow is now more dependable.

Payments and production reliability moved from setup to functional rollout.

• Payments are live: Added paid plans and credit purchase flows for users.
• Sign-in fix: Resolved a Google sign-in issue that affected some production users.
• Reliability improvements: Fixed several production issues to improve overall stability.
• Legacy branding cleanup: Removed lingering legacy naming in UI/docs and standardized branding.

This release focused on improving security for critical user actions.

• Stronger protections: Added additional safeguards around sensitive platform operations.
• Safer event handling: Improved how background events are validated and processed.
• Improved request checks: Tightened validation for account and assistant-related actions.

This is the first public foundation of the product.

• Core product launch: Released assistants dashboard, authentication, and essential user flows.
• Assistant lifecycle support: Enabled creating, managing, and operating assistants end to end.
• Foundational docs and tools: Shipped initial docs and setup tooling.

Every assistant inbox now has a handle you control. Instead of the auto-generated address you got when email was first enabled, you can pick something short and memorable and hand it out with confidence. The change applies to both OpenClaw and Hermes assistants.

• Choose your own handle: On the Email tab for any assistant, type the name you want before the @ and save. Your inbox address updates immediately and any future email sent by the assistant goes out from the new address.
• Change it whenever you want: Renamed your assistant, or just want something catchier? Update the handle again and the new address takes effect right away. No redeploy to babysit on OpenClaw, and a seamless redeploy on Hermes.
• Globally unique, guaranteed: Handles are reserved the instant you claim one, so two assistants can never share an inbox. If someone else has already taken the name you want, you'll see a clear message and can try another.
• Friendly validation as you type: The form tells you right away if a handle is too short, uses unsupported characters, or touches one of the reserved mailbox names. No guessing, no trial-and-error saves.
• Defaults unchanged if you don't care: Assistants that already have email enabled keep their existing address — nothing moves until you choose a new handle.
• A cleaner delete confirmation: Deleting an assistant now opens a proper in-app dialog instead of a browser prompt, with a loading state while the tear-down runs and a smooth exit animation on the assistants list.

A random identifier is fine for an internal ID, but it's a terrible thing to hand out to colleagues, customers, or vendors. Now your assistant's inbox can look like it belongs to a person, not a database row.

Hermes is now more than 50% faster. Below is time to first reply — how long between a user's message and the assistant's first word back — tracked over the last few weeks. The baseline held steady for a long time; this release is the cliff on the right.

Two shifts landed in this release. Replies are dramatically faster (above), and assistants now keep their full state across redeploys — every conversation, uploaded file, scheduled job, and approved pairing survives.

• Over 50% faster replies: Simple questions that used to take about 30 seconds now come back in around 7. Complex tasks that took a full minute now finish in roughly 20.
• Automatic snapshots on redeploy: The moment you trigger a redeploy, your assistant's state is captured and saved before the running instance is replaced. Nothing extra to configure, nothing to remember.
• Seamless restore on the new instance: When the replacement instance comes up, it boots with the exact workspace, memory, and channel pairings the previous one had. Conversations pick up right where they left off.
• Existing assistants migrated automatically: Hermes assistants created before this change have their prior data moved over into the new snapshot format, so long-running assistants keep their history without any manual work.
• More forgiving redeploys: A redeploy no longer fails just because your assistant is busy working through a backlog or waiting on an external provider. The platform gives it room to settle instead of restart-looping.
• Faster channel reconnection: After a redeploy, Telegram pairings and other approved channels are already restored, so there's no re-pairing or re-authorization step.

Redeploys used to be the scariest button in the product for long-running Hermes assistants, and waiting on replies was the slowest part of a typical day. With this release, shipping an update is routine instead of a data-loss risk, and your assistant answers while you're still looking at the screen.

When you create a managed assistant, the model picker now pulls in the full OpenRouter catalog instead of a short curated list. That means you can browse and search across 500+ models from dozens of providers, all in the same dropdown. We also updated the defaults so new assistants start on the latest models.

• Full model catalog: The model selector during onboarding now shows every model available through OpenRouter, organized by provider. Popular providers like Anthropic, OpenAI, Google, and DeepSeek appear first, and you can type to search the rest.
• Instant search across all models: Start typing a model name, provider, or ID and the list filters down immediately. No more scrolling through a fixed list hoping the model you want is there.
• GPT-5.4 Mini as the new default: Managed assistants now default to OpenAI's GPT-5.4 Mini, which offers strong performance at a lower price point. You can always change it before confirming.
• GPT-5.4 and GPT-5.4 Mini added: Both models are now available for managed and bring-your-own-key assistants, so you can use them regardless of how your assistant is set up.
• Context length at a glance: Each model in the dropdown shows its context window size, so you can make an informed choice without leaving the setup flow.

This update makes it much easier to find and use the right model for your assistant, whether you want the latest flagship or a cost-effective option for simpler tasks.

Hermes assistants now support inbound email. When someone sends a message to your assistant's email address, it arrives as a structured webhook event that the agent can process, respond to, and take action on. This brings Hermes to parity with OpenClaw on one of the most requested messaging channels.

• Email forwarding for Hermes: Inbound emails are routed to Hermes assistants through the same address-based forwarding that OpenClaw assistants already use, so setup works the same way.
• Structured webhook delivery: Hermes receives emails as typed webhook events with the sender, subject, body, and message threading information preserved, so the agent has full context to work with.
• Persistent storage for Hermes containers: Hermes assistants can now use durable storage that survives container restarts, so files, state, and workspace data stick around between sessions.
• More reliable terminal startup: The health sidecar now lazy-loads the terminal emulator, avoiding startup failures when the native module isn't immediately available inside the container.

Email is one of the simplest ways to interact with an assistant without installing anything. With this update, both OpenClaw and Hermes assistants can be reached the same way.

Hermes assistants now come with a built-in terminal in the dashboard. That means you can open a live shell directly inside the assistant's container, inspect files, run commands, and debug issues in place instead of guessing from the outside.

• A dedicated Terminal tab: Active Hermes assistants now have their own Terminal tab on the assistant details page, so shell access is part of the normal management flow instead of a separate workaround.
• Live shell access inside the container: You can connect to the running Hermes environment, execute commands, inspect the filesystem, and verify what is actually happening on the deployed assistant.
• Faster debugging when something feels off: If an assistant is behaving unexpectedly, you now have a direct way to check logs, files, processes, and runtime state without waiting for a redeploy or trying to infer everything from the UI.
• Secure, short-lived terminal sessions: Terminal access is gated behind signed session tickets and tied to the active assistant session, so it is designed for controlled troubleshooting rather than permanent open access.
• A more dependable Hermes runtime underneath: We also tightened the runtime path behind the scenes so terminal sessions connect more reliably and deploy status updates are less likely to fail during transient errors.

This is the most direct way we have added yet to work with a hosted assistant like a real running system, not just a dashboard object. For Hermes users especially, it makes day-to-day debugging and inspection much more practical.

getclaw now supports Hermes Agent as a hosted framework. That means you can create a Hermes assistant directly from the product, deploy it through the same managed flow, and operate it from the same dashboard you already use for OpenClaw assistants.

• A second hosted framework: You can now choose Hermes Agent instead of OpenClaw when creating a new assistant, directly inside getclaw.
• Managed through the same dashboard: Hermes assistants live inside the same deployment, management, and post-launch workflow as the rest of the platform, so there is no separate hosting path to learn.
• Telegram support for Hermes deployments: Hermes assistants can now run on Telegram in the hosted environment, so messaging works as part of the normal getclaw channel setup flow.
• Approval controls for Telegram access: When a restricted user messages a Hermes assistant, their Telegram user ID shows up in the dashboard so you can approve them without leaving getclaw.
• A more usable hosted Hermes runtime: Hermes deployments now boot cleanly in the managed environment, which makes hosted Hermes much closer to a real day-to-day option instead of an experimental add-on.

This is a framework expansion, not just a new toggle. getclaw now supports two different open-source agent runtimes, which gives users a clearer choice between OpenClaw's operations-first model and Hermes's more personal, persistent-agent model.

Until now, each conversation with your assistant started from a blank slate. The assistant could be brilliant in the moment, but the next time you talked to it, all of that context was gone. Lossless-claw changes that. It gives every managed assistant a persistent, structured memory that grows over time — so your assistant gets more useful the more you use it.

• Full conversation memory: Lossless-claw captures the complete context of every conversation — not just summaries or keywords, but the actual substance of what was discussed, decided, and learned.
• Automatic context recall: When your assistant starts a new conversation, it pulls in relevant context from past interactions. It knows what you've worked on, what you've asked before, and what preferences you've expressed — without you having to repeat yourself.
• Gets smarter over time: The more you use your assistant, the better it understands your projects, your terminology, and how you like things done. This is the difference between a tool that resets every session and one that actually learns.
• On by default, no setup: Every assistant now ships with lossless-claw pre-installed and active. There is nothing to configure. If you need to disable it for a specific assistant, set the DISABLE_LOSSLESS_CLAW environment variable.
• Baked into the image: The plugin is pre-installed in the container image at build time, so there is no download delay on cold starts and no risk of install failures at runtime.

This is a step change in what managed assistants can do. An assistant that remembers is fundamentally more useful than one that forgets — for ongoing projects, repeated workflows, and anything where continuity matters.

The infrastructure underneath managed assistants is now faster to start, easier to debug, and lighter on resources. This release removes unused dependencies from the container image, parallelizes data restores, and finally gives you visibility into what actually went wrong when a deploy fails.

• A smaller, faster container image: Playwright and pnpm have been removed from the Docker image. The result is a leaner image and faster startup.
• Parallel R2 restores: Workspace and skills data now restore from R2 concurrently instead of sequentially, cutting restore time on restarts.
• Delete backup data from the dashboard: You can now delete individual files or entire directories from your assistant's R2 backup storage directly in the dashboard, with streamed progress for large deletions.
• Safer Telegram defaults: If a Telegram allowlist is configured with no approved users, the runtime now falls back to pairing mode automatically instead of rejecting all messages.

Assistants start faster, and when something breaks during startup, you can actually see why.

Managing channels is now much cleaner, and the runtime underneath managed assistants just took a meaningful step forward too. This release adds a dedicated Messaging tab for channel setup and rolls hosted assistants onto OpenClaw 2026.3.13, which pulls in the biggest recent capability upgrades from upstream.

• A dedicated Messaging tab: Telegram, Slack, Discord, and Email now live in one place on the assistant details page, so channel setup no longer feels scattered across separate screens and controls.
• Easier channel management: You can enable or disable channels, rotate bot tokens, see whether a channel is configured, copy an assistant's email address, and jump straight into email management without leaving the dashboard.
• The big OpenClaw upgrade is now in managed deployments: Moving to OpenClaw 2026.3.13 brings the biggest recent upstream improvements into the hosted runtime, including browser automation, assistant email support, and unlimited built-in search.
• A safer and more predictable runtime: Deployments now validate the OpenClaw config before startup, set Control UI origins more explicitly for hosted environments, and pin Telegram streaming to a safer default so upgrades are less likely to break in surprising ways.

This is partly a dashboard release and partly an infrastructure one, but both sides matter. Channel operations are easier day to day, and managed assistants are now much closer to the current OpenClaw experience instead of lagging behind it.

Search is now available to every OpenClaw user, and it ships with unlimited usage from day one. No plan upgrade, extra toggle, or special setup required. If you already have assistants deployed, just redeploy each instance once to enable the new feature.

• Search is now built in: OpenClaw assistants can now use search as a native capability, opening up research and retrieval workflows that were not available before.
• Unlimited usage included: Every user gets the feature with no search cap, so assistants can keep using it without running into usage limits.
• One redeploy to enable it: Existing instances need a redeploy to pick up the new runtime behavior. Once redeployed, the feature is active automatically.
• Applies to your current assistants: This is not limited to newly created assistants. Your existing instances can use search too as soon as they are redeployed.

This is a meaningful capability release, not just a limit change. Search is now part of what every assistant can do, and if you already have one live, a redeploy is all it takes to turn it on.

OpenClaw assistants now have their own email address. They can check their inbox, send messages, and reply to threads — all as part of a workflow.

• Dedicated email address: Each assistant gets its own email address, provisioned automatically when the skill is enabled. No external setup required.
• Send, reply, and check inbox: The assistant can compose new emails, reply within existing threads, and read incoming messages on its own.
• Generous free tier: Every user gets 100 free emails per month. Beyond that, additional emails cost fractions of a cent from your credit balance.
• Threaded conversations: Replies are linked to the original message, so recipients see a clean conversation thread instead of disconnected messages.

Email turns your assistant from something you talk to into something that can talk to others on your behalf — following up, reporting, and communicating without you in the loop.

OpenClaw can now use a real browser as part of its workflow. That means your assistant is no longer limited to APIs and text-only tools. It can move through the web the way a person would: opening pages, clicking buttons, reading content, and acting on what it finds.

• Navigate real websites: OpenClaw can visit live pages, wait for interfaces to load, and work through multi-step browser flows instead of stopping at static HTML.
• Interact with the page: Clicking, form filling, scrolling, and page-to-page navigation are now part of the assistant's toolkit, which unlocks workflows that were previously impossible with API-only access.
• Read what users actually see: The assistant can inspect rendered pages, not just backend responses, which is a big step forward for research, operations, QA, and support-style tasks.
• Unlock broader automation: This creates room for assistants that verify dashboards, collect information from tools without public APIs, reproduce UI issues, and complete browser-based tasks end to end.

This is one of the biggest capability jumps in the product so far. Giving OpenClaw a browser turns it from an assistant that can call tools into one that can operate directly on the web.

This release is about polish where it matters: assistants start more smoothly, the first-run experience feels more alive, and Telegram access is less likely to surprise you later.

• More dependable launches: Deployments now recover more gracefully during slower startup moments, so assistants are less likely to get stuck while going live.
• Telegram access that stays put: If you maintain a private Telegram allowlist, those approved users now stay approved more reliably across edits and redeploys.
• Clearer save and redeploy feedback: When you update files from the dashboard, you now get much better progress messaging so it's obvious what was saved and what's still rolling out.
• A more premium first impression: Persona previews load faster, hover demos feel smoother, and the homepage greeting feels more intentional instead of generic.

Small release on paper, but it removes a lot of friction in the moments users actually notice: launch, preview, and trust.

Creating an assistant no longer means filling out forms. Just describe what you want in a conversation, and the wizard configures everything for you.

• Chat-based creation: Describe your assistant in plain language — name, personality, model, API key, and channels are all set through a guided conversation.
• 6 pre-built personas: Get started instantly with LinkedIn Influencer, News Reporter, Sales Rep, Customer Support, Software Developer, or Personal Coach — each with a ready-to-go personality and skills.
• Redesigned homepage: A new homepage showcases what's possible with persona demos and video walkthroughs.
• Fresh visual identity: Updated logo, favicons, and branding across the product.

Two improvements to help you understand and manage running assistants faster.

• Live server logs: A new logs panel on the assistant detail page streams your worker's output in real time, with color-coded log levels and auto-scrolling.
• Playback controls: Start, stop, and clear the log stream without leaving the dashboard.
• Faster config changes: Updating your model or API key now takes effect in about 30 seconds, down from a full redeploy that could take several minutes.
• Visual update status: The dashboard shows an "Updating configuration..." indicator while changes are being applied.

Your assistant is no longer limited to a single messaging platform. You can now deploy to Telegram, Slack, Discord, or any combination of the three.

• Slack support: Connect your assistant to a Slack workspace using a Bot Token and App Token. A built-in setup guide walks you through creating the Slack app.
• Discord support: Connect your assistant to a Discord server with a single bot token. The setup guide covers bot creation, permissions, and server invitation.
• Multi-channel deployments: Enable multiple channels on a single assistant. The detail page shows all active channels at a glance.
• Step-by-step setup guides: Each channel includes an expandable guide with the exact steps, scopes, and permissions needed.

Until now, editing assistant files meant working through openclaw UI or redeploying from scratch. The new file editor gives you direct access to your assistant's files right from the dashboard.

• Browse your assistant's files: Navigate directories, open files, and explore your assistant's configuration and workspace from a familiar file explorer.
• Edit with syntax highlighting: A full-featured code editor with automatic language detection for JSON, TypeScript, JavaScript, YAML, Markdown, shell scripts, and more.
• Save & redeploy in one click: Edit a file and redeploy your assistant in a single action, no need to leave the dashboard.
• Safe by default: Unsaved changes are flagged before closing, and saves are validated to prevent conflicts.

This update improves how managed assistants preserve state, so restarts are less disruptive and recovery is more dependable.

• Better data safety: Backups now run more consistently, including periodic full backups to reduce the chance of drift over time.
• Smoother restarts: Assistants now recover newer saved state more reliably when containers restart.
• More complete recovery: Restores now cover assistant state, workspace files, and installed skills.
• Clear backup health: Admin status now shows last successful backup, last backup attempt, and recent backup errors when present.
• Customer value: Less manual recovery work, fewer surprises after restarts, and more confidence that assistant progress is preserved.

We made this change because cloud browser automation setups were too fragile and required too much manual configuration. The goal of this release is to give you safer defaults, fewer deployment steps, and more reliable assistant behavior in isolated environments.

• Why this matters: Cloud environments can fail in ways that are hard to debug when browser connectivity is not configured securely and consistently.
• What you now get: Deployments now apply secure connectivity defaults automatically, so assistants come up with fewer manual steps.
• Reliability improvement: Isolated sandboxes now default to the Playwright CLI browser path, which avoids common extension-relay failures.
• Stability improvement: Gateway configuration is normalized for current OpenClaw runtimes, reducing startup and connectivity issues.
• Customer value: Faster time-to-working assistant, less deployment troubleshooting, and safer browser automation defaults out of the box.

We are introducing first-class support for fully customizing your openclaw.json, alongside making assistant configuration easier to understand and more dependable.

• Customize openclaw.json: You now have direct control to customize your openclaw.json configuration just how you want it.
• Clearer config status: You can now quickly see which settings are currently applied versus still pending.
• Easier troubleshooting: Side-by-side config comparisons make it faster to spot what changed.
• More reliable updates: Config saves are now applied more consistently across assistant redeploys.
• Cleaner comparisons: Runtime-managed values are filtered out so you can focus on user-controlled changes.
• Supported OpenClaw version: 2026.2.19.

This release focused on product clarity and assistant configuration flow.

• Model updates are easier: Change your assistant model directly from the assistant details page.
• Cleaner logged-in experience: Returning users are now guided to their assistants right away.
• Improved onboarding content: New comparison content helps users choose the right path faster.

This update focused on clarity and usability.

• Smoother account flows: Improved key flows so setup and activation feel more reliable.
• Better docs navigation: Updated structure and links so it is easier to find what you need.
• Clearer guides: Expanded and refined help content for common usage patterns.

This release made external integrations more complete while improving day-to-day stability.

• More complete integrations: Added broader support for assistant management tasks.
• Easier setup: Updated guides and behavior for a smoother integration experience.
• Stability improvements: Fixed issues that could affect account, billing, and assistant operations.

This patch focused on making managed usage more dependable.

• More accurate credits: Resolved edge cases around balance updates and top-ups.
• More stable assistant performance: Fixed issues that could interrupt managed assistant responses.

A full pass on the landing experience improved readability and conversion-oriented messaging.

• New page structure: Improved layout and section flow to make the value proposition clearer.
• Visual and copy polish: Updated visuals and messaging to improve readability and conversion.

This was a major milestone for managed assistant billing and reliability.

• Managed credits are live: Users can now run assistants with platform-managed credits.
• Improved account isolation: Credits are now separated per user for better consistency.
• Updated pricing structure: Managed assistant plans were clarified and separated from other offerings.
• Deployment reliability improvements: Managed assistant deployment flow is now more dependable.

Payments and production reliability moved from setup to functional rollout.

• Payments are live: Added paid plans and credit purchase flows for users.
• Sign-in fix: Resolved a Google sign-in issue that affected some production users.
• Reliability improvements: Fixed several production issues to improve overall stability.
• Legacy branding cleanup: Removed lingering legacy naming in UI/docs and standardized branding.

This release focused on improving security for critical user actions.

• Stronger protections: Added additional safeguards around sensitive platform operations.
• Safer event handling: Improved how background events are validated and processed.
• Improved request checks: Tightened validation for account and assistant-related actions.

This is the first public foundation of the product.

• Core product launch: Released assistants dashboard, authentication, and essential user flows.
• Assistant lifecycle support: Enabled creating, managing, and operating assistants end to end.
• Foundational docs and tools: Shipped initial docs and setup tooling.