How can we help?
Browse what is written down about getting started, accounts, agents, billing and security, then search it.
Still need a person?
Contact the team when the answer is not written down here.
Browse the full workspace capability list
Help articles
Getting started
- What is Ethen?An overview of the Ethen AI workspace platform and what you can do with it.
- Creating your first workspaceOpen the console, type in Cortex, or launch a specialist agent workspace from Fleet.
- Navigating the sidebarHow to use the sidebar to switch between agents, sessions, and settings.
- Using quick actionsLaunch Code, Browser, Design, and other workspaces directly from the home screen.
- Mock vs. live agents — what's the difference?Mock agents simulate responses locally. Live agents connect to real provider APIs.
Account and billing
- Signing in to EthenEthen uses Clerk for authentication. Sign in with your email or SSO provider.
- Understanding your credit balanceCredits are consumed per agent request. View your balance in Settings → General.
- Billing and plansBilling plans and payment processing require a live billing backend. Currently in prototype.
- How to log outOpen your account menu at the bottom of the sidebar and select Log out.
- Changing your display name or avatarProfile details are managed through Clerk. Open Settings → Profile to manage your account.
Chats and sessions
- Starting a new conversationClick the Ethen logo or use ⌘K to open a new workspace or start a chat.
- Viewing recent sessionsYour most recent sessions appear at the top of the sidebar for quick access.
- Session persistenceSessions are stored in your browser by default. Durable storage requires Supabase configuration.
- Sharing a conversationConversation sharing is not yet available. It is planned for a future update.
- Deleting chat historyYou can reset local preferences in Settings → Data Controls → Actions.
Models and usage
- Which models does Ethen support?Ethen supports OpenAI, Anthropic, and any OpenAI-compatible endpoint.
- Configuring a model providerGo to Settings → Models & Providers to add an API key for OpenAI or Anthropic.
- Switching models in a conversationUse the model switcher in the composer bar to change the active model mid-session.
- Token limits and context windowsDefault context window is 32K tokens per conversation. Larger windows consume more credits.
- Usage trackingUsage events are recorded in Supabase when configured. View totals in Settings → Billing & Credits.
Voice and microphone
- Enabling voice inputVoice dictation requires an OpenAI API key set in your environment (OPENAI_API_KEY).
- Using the microphone buttonClick the microphone icon in the composer to start voice dictation when configured.
- Microphone permissionsEthen requires browser microphone permission. Grant access when prompted by your browser.
- Text-to-speech / read aloudText-to-speech playback is planned. It is not available in the current build.
- Voice transcription rate limitsTranscription requests are rate-limited per minute based on your API plan.
Agents and Fleet
- What is Fleet?Fleet is your agent roster — a browsable list of all specialist agents available in Ethen.
- Launching an agentClick any agent in Fleet to open its workspace. Mock agents work without any setup.
- Agent trust states explainedEach agent has a trust state: Live, Preview, Mock, Setup Required, Planned, and more.
- Setup-required agentsThese agents need provider credentials. Configure them in Settings → Models & Providers.
- What happens in mock mode?Mock agents return simulated responses. No real external systems are affected.
Flow and connected apps
- What is Flow?Flow lets you chain steps, connect apps, and automate tasks.
- Connecting Google WorkspaceDrive, Docs, Sheets, Gmail, and Calendar integration requires OAuth configuration.
- Connecting Microsoft 365Outlook, Teams, OneDrive, and SharePoint require Microsoft OAuth setup.
- GitHub integrationConnect GitHub for repository and issue access via OAuth in Connected Apps settings.
- Connecting SlackSlack channel and message integration requires a Slack app installation.
Studio and media tools
- What is Studio?Studio is Ethen's media workspace for image generation, video, audio, and editing.
- Generating imagesImage generation requires an image model provider (e.g., DALL-E via OpenAI API key).
- Storage for media assetsAssets use in-memory storage by default and are cleared on server restart without Supabase.
- Supported media formatsStudio supports common image, video, and audio formats depending on the configured provider.
- Studio is in early accessMedia tools are early-stage. Some features may be unavailable or limited in scope.
Security
- How Ethen protects your API keysAPI keys are never displayed in the UI and are redacted from logs, traces, and audit records.
- Tool approval and risk levelsTools are classified as low, medium, or high risk. High-risk tools require confirmation every time.
- Audit logEthen records tool executions, approvals, and credential operations in an in-memory audit store.
- Exporting audit eventsYou can export the audit log as JSON or CSV from Settings → Security & Approvals.
- Encryption and data at restConnected-app credentials are stored encrypted in a managed secrets vault and used only by Ethen's servers. See the Enterprise Security document.
Troubleshooting
- Agent is not respondingCheck that your provider API key is configured and valid in Settings → Models & Providers.
- Session data disappearedSessions stored in localStorage are cleared on cache clear or incognito mode. Configure Supabase for persistence.
- Voice input not workingEnsure OPENAI_API_KEY is set and your browser has microphone permission granted.
- Build or type errorsRun pnpm typecheck and pnpm lint from the project root to diagnose issues.
- Resetting preferencesGo to Settings → Data Controls → Actions and click Reset preferences to clear local settings.
Workspace capabilities
Where this list comes from
Workspace capabilities below are read from bundled seed data and the trust-state registry, not from a live availability feed. A trust label describes the fixture, never a certified live service.
- ResearchPreview
Preview web search, news search, keyword research, SEO analysis, and source discovery.
Workspace available, but some behavior depends on preview data or scoped integrations.
- Web search and news aggregation are powered by preview search APIs — result freshness and coverage depend on upstream providers.
- No access to authenticated research databases, paywalled journals, or premium data sources.
- SEO and keyword analysis are reference-only — no real search-engine data is available.
- Generated research briefs should be fact-checked before use.
- StudioMock
Create, edit, animate, voice, and manage AI media from one Ethen workspace.
Demo mode — responses are simulated, no live backend.
- Live media generation requires a provider API key — currently in mock preview mode.
- No real video editing, audio production, or canvas manipulation is implemented.
- Generated media are placeholder directions only, not production assets.
- Asset storage and project persistence are browser-local only in this repo.
- AcademiaMock
Research papers, citations, literature reviews, and scholarly analysis.
Demo mode — responses are simulated, no live backend.
- Paper discovery and citation generation are mock/preview — no live academic database or API is connected.
- Literature reviews are AI-generated and should be verified against original sources.
- No direct access to paywalled journals, PDF repositories, or preprint servers.
- Citation formatting may not match all style-guide edge cases — always verify generated references.
- ChatPreview
A full-featured general-purpose chatbot with branching, editing, and response controls.
Workspace available, but some behavior depends on preview data or scoped integrations.
- General-purpose chat agent — no specialized tool access or workflow integrations are available.
- Responses are generated by AI models in preview mode; verify critical information before acting on it.
- No persistent memory across sessions beyond the current conversation thread.
- No access to user files, external APIs, or live data sources in this workspace.
- GlobeMock
Interactive globe explorer with places, layers, saved views, and geospatial intelligence.
Demo mode — responses are simulated, no live backend.
- Mock preview mode — no real satellite imagery, terrain data, or geocoding API is connected.
- Geospatial briefs are AI-generated and should be verified against authoritative sources.
- Saved views, bookmarks, and search history are browser-local only in this preview.
- Real-time location data, directions, and transit information are not available.
- ComputeMock
Use a controlled browser/computer workspace for testing, browsing, and workflow tasks.
Demo mode — responses are simulated, no live backend.
- Static workspace shell — no real browser automation, Playwright, or backend runtime is connected.
- All UI states (running, approval, paused, takeover) are mock/local-only previews.
- No real screenshots, viewport streaming, or browser sessions are implemented.
- Approval flow, permission scoping, and action timeline are visual mockups only.
- FounderPreview
Build your company from idea to launch with structured workstreams.
Workspace available, but some behavior depends on preview data or scoped integrations.
- Business briefs, market research, and website drafts use mock/preview data — no live providers are connected.
- Content & copywriting handoffs route to the Writing Agent workspace.
- Brand & screen design handoffs route to the Designer workspace.
- Legal setup and documents are marked as manual-review / not-available.
- No incorporation, filing, banking, or real company formation is implemented.
- Universal ComposerPreview
Start from the shared Ethen composer for chat, prompts, and cross-workspace handoff.
Workspace available, but some behavior depends on preview data or scoped integrations.
- This card opens the shared composer surface only — it does not add new agent capabilities by itself.
- Any app access, workflow execution, or write actions still depend on the destination workspace and its readiness state.
- Connector setup, approvals, and dry-run constraints are enforced in the target workspace, not here.
- FlowPreview
Build, draft, simulate, and review workflow automations across apps.
Workspace available, but some behavior depends on preview data or scoped integrations.
- Builds draft, simulation, and review flows only — it does not perform live execution or activation from the marketplace entry.
- Connected app coverage and route status depend on the current preview workflow catalog in this repo.
- Approvals, simulations, and readiness signals should be reviewed before relying on a workflow outside preview use.
- SecurityPreview
Run read-only security review setup, scan planning, and evidence review for authorized repositories.
Workspace available, but some behavior depends on preview data or scoped integrations.
- Authorized repositories only — no live target scanning, credential testing, exploit automation, pull requests, or disclosure actions.
- Sentinel focuses on planning, static analysis, and evidence review in this release.
- Human review is still required for approvals, disclosures, and any remediation outside preview flows.
- CodingBeta
Write, review, and debug code with an AI pair programmer.
Feature-complete but still in active development — behavior may change.
- Preview workspace — no live code execution, deployment, or repository write access from this entry.
- Human review is required before applying any AI-suggested code changes to production.