Shi Yu
Morning. Review this week's new registrations, draft personalized openers for the qualified leads, and help me book some demos.
12:38 Read 2/2 ✓✓FAQ
Answers about authorization, third-party platforms, product usage, security, and privacy.
The world’s first workspacefor AI-Native teams
I. Cloud Agent Authorization
What is Cloud Agent Authorization?
Cloud Agent Authorization grants an AI employee running on a remote server the minimum permissions required to operate third-party platforms on your behalf.
Cloud host: a remote server without a physical screen or local browser
Agent: an AI program that can execute tasks autonomously
Authorization: permission to work with services such as GitHub, Lark, Telegram, and cloud APIs
Q1: What permissions does the Agent need?
OpenMax follows the principle of least privilege and requests only the access needed for the task you approve.
Permission Scope
| Task Type | Required Permission | Scope |
|---|---|---|
| View server status | Read-only | Cannot modify configuration |
| Deploy or update apps | Specific directory write access | Limited to approved paths |
| Restart services | Service management | Limited to approved services |
| View logs | Log read access | Excludes sensitive configuration |
Irreversible actions require confirmation. Credentials are encrypted, third-party tools are reviewed, and operations are logged for audit.
Q2: Which authorization method should I use?
Authorization Methods
| Method | Best For | Security Guidance |
|---|---|---|
| SSH Key | Linux servers and VPS management | Use a dedicated account and restrict paths |
| API Token | Cloud platforms and SaaS APIs | Use scoped, expiring, dedicated tokens |
| Browser Login | Web-only workflows and MFA | Use the isolated virtual browser and clear sessions |
For SSH, create a dedicated non-root account and add only the approved public key.
For APIs, create a dedicated fine-grained token with an expiry date.
For browser workflows, install the Browser component and log in through the isolated VNC session.
Q3: How does authorization work on a headless cloud host?
A cloud server does not need a physical desktop. Use a Personal Access Token for API-based services, or install the Browser component and connect through VNC for full browser authentication.
Plaintext
Help me configure the minimum GitHub permissions required to create and merge pull requests.Prefer scoped Personal Access Tokens over sharing cookies or personal session credentials.
Q4: How do I revoke authorization?
Tell your AI employee which platform authorization to revoke.
Delete the SSH key, API token, or connected app on the source platform.
Clear the virtual browser session when browser login was used.
Confirm that the integration can no longer perform new operations.
Q5: Are credentials retained after service ends?
After you disconnect an integration, OpenMax deletes the stored credentials and stops initiating operations. For complete revocation, also remove the corresponding credential on the source platform.
II. Social Media & Third-Party Platforms
Q6: Which third-party platforms are supported?
OpenMax can connect to supported APIs and use the Browser component for web-based platforms. Availability depends on the platform's integration options and your approved use case.
Plaintext
Can you help me operate this platform, and what is the safest integration method?Q7: How is data timeliness maintained?
Scheduled collection and proactive notifications at your configured frequency
Incremental updates that process only new content
Failure alerts for timeouts, rate limits, and unavailable sources
Data Latency Factors
| Factor | Impact | Description |
|---|---|---|
| Collection frequency | Configured delay | Depends on the interval selected for the task |
| Platform API rate limits | Occasional delay | Some platforms restrict request frequency |
| Viral content spikes | One or two cycles | Rapidly changing content may be discovered later |
Q8: Is public data collection compliant?
Collect only publicly accessible content
Respect platform terms, robots rules, and rate limits
Do not collect private messages or restricted content
Use collected data according to applicable privacy and data-protection laws
III. Product Usage
Q9: How is OpenMax different from a standard AI chat app?
Product Comparison
| Dimension | Standard AI Chat | OpenMax |
|---|---|---|
| How you use it | Open the app and ask questions | Assign tasks and receive completed results |
| Memory | Mostly conversation-bound | Persistent cross-conversation memory |
| Automation | Manual prompts | Scheduled and proactive task execution |
| Channels | Dedicated chat app | Lark, Telegram, Zalo, Teams, and more |
| Collaboration | Usually personal | Shared AI employee in team conversations |
Q10: How do I get started?
Register an OpenMax account and choose a plan.
Follow the Channel Deployment Guide to connect a messaging platform.
Select roles and complete onboarding in Dashboard.
Send your first message and begin collaborating.
For deployment assistance, contact contact@openmax.com or join the OpenMax Telegram community.
Q11: Which AI models does OpenMax support?
OpenMax is a fully managed service. Model selection and orchestration are handled by the platform, so users do not need to provide an API key.
Model Tiers
| Plan | Model Tier | Capabilities |
|---|---|---|
| Air | Mainstream standard model | Writing, organization, and everyday analysis |
| Pro / Ultra | Top-tier reasoning model | Complex reasoning, long documents, and high-frequency work |
IV. Security & Privacy
Q12: How does OpenMax protect data and privacy?
HTTPS/TLS in transit and encrypted storage
Conversation data is not used for model training
Strict access controls and audit logs
No personnel access without explicit user authorization
User data is not sold or shared for unrelated purposes
Use dedicated credentials, grant only the permissions required for each task, and revoke access when an integration is no longer needed.