FAQ & Support

This page summarizes the most common user questions and solutions. If your question is not listed here, please contact us through the support channels below.

1. Account

1.1 How to register an account?

Visit www.octoreport.com and click the "Get Started" button:

• Email + Password: Fill in your email and password (8+ characters, including letters and numbers), verify your email to complete registration
• Magic Link: Enter your email, the system will send a login link to your email, click to login (no password needed)
• Third-party Login: Supports quick login with Google and GitHub accounts

💡New users automatically receive 10,000 credits after successful registration, enough to complete multiple data collections and report generations.

1.2 What if I forget my password?

If you forget your password, there are two solutions:

1. Use Magic Link Login (recommended): Select "Magic Link" on the login page, enter your registered email, and click the link in the email to login
2. Reset Password: Click the "Forgot Password" link on the login page, enter your registered email, and follow the email instructions to reset your password

1.3 Can I change my email?

Currently, direct email change is not supported. If you have special needs, please contact customer support.

1.4 How to delete my account?

Go to the "Settings" page and find the "Delete Account" button at the bottom:

• After deleting your account, all data (sources, libraries, reports, Ask history) will be permanently deleted
• Unused credits will be cleared and cannot be refunded
• This operation is irreversible, please proceed with caution

⚠️Important Reminder: Before deleting your account, please ensure you have exported or backed up important data.

2. Feature Usage

2.1 Data source created but not automatically collecting content?

Please check the following:

• Schedule Strategy Settings: Confirm that the data source schedule strategy is not "Manual Trigger". If it's interval mode or weekly plan mode, check if the "Next Execution Time" has arrived
• Sufficient Balance: Check if the account has enough credits (each collection requires 1-50 credits)
• Task Logs: Go to the "Task Logs" page to see if there are any failed records, and if so, check the failure reason
• Manual Trigger: You can click the "Execute Now" button on the data source details page to manually trigger a collection

2.2 RSS data source returns 403 or 404 error?

This is usually because the target website has access restrictions or RSSHub routing configuration issues:

• 403 Forbidden: Website requires login or Cookie to access
  • Go to data source configuration and enable "Cookie Configuration"
  • Refer to RSS Feed Documentation to configure global or custom Cookies
• 404 Not Found: RSSHub route does not exist or path is incorrect
  • Use the "RSSHub Route Search" feature to find the correct route
  • Check if route parameters are filled in correctly (e.g., user ID, category, etc.)
• RSSHub Instance Failure: The system will automatically switch to a backup instance, retry later

2.3 Report generation failed, how to troubleshoot?

Common causes and solutions:

2.4 Ask cannot find relevant content?

Ask's library mode is based on keyword retrieval, if content cannot be found:

• Check Library: Confirm that the selected library has relevant content
• Adjust Question Phrasing: Use more specific keywords, for example:
  • ❌ Poor question: "Any news?"
  • ✅ Good question: "What news about AI large language models in the past 3 days?"
• Switch to General Mode: If library mode cannot meet your needs, you can switch to general mode (not based on library)

2.5 Trigger Inbox (email channel) did not receive a reply?

Please check the following:

• Email Whitelist: Confirm that the sending email has been added to the "Email Whitelist" (Settings page)
• Email Subject and Content: The email subject should include the topic, and the body should clearly describe the time range and requirements
• Credit Balance: Each Trigger Inbox request on the email channel requires about 100 credits. Check whether the balance is sufficient
• Daily Quota: By default, a maximum of 10 Trigger Inbox requests per day is allowed. Requests above the quota will be rejected
• Spam Folder: Check if the reply email has been mistakenly identified as spam

💡Tip: Trigger Inbox requests on the email channel send progress notifications. If you do not receive any reply (including progress notifications) within 5 minutes, the request may have been rejected. Please check the whitelist configuration.

3. Billing & Credits

3.1 How to obtain credits?

Currently there are two ways to obtain credits:

• New User Gift: Automatically receive 10,000 credits after successful registration
• Redemption Code: Enter a redemption code on the "Credit Management" page to recharge (redemption codes can be obtained through activities or purchases)

Online payment recharge will be supported in the future.

3.2 How are credits consumed?

The credit cost for different operations is as follows:

3.3 Will failed tasks be charged?

No. OctoReport uses an "atomic billing" mechanism:

• Before the task starts, the estimated cost is deducted
• After the task is completed, the actual cost is calculated
  • If successful → settled according to actual cost, excess credits refunded
  • If failed → full refund, no credit loss
• All charge and refund records can be viewed on the "Credit Management" page

✅System Guarantee: Database transactions ensure the atomicity of charges and execution, there will never be a situation where "charged but task not executed" or "executed but not charged". See Atomic Billing technical documentation.

3.4 How to view credit consumption records?

Go to the "Credit Management" page to view:

• Current Balance: Account available credits
• Consumption Details: All charge and refund records, including operation type, time, amount, and status
• Filter Function: Filter by operation type (collection, cleaning, report, Ask) and time range

3.5 Why are there negative numbers (refunds) in consumption details?

Negative numbers indicate refund operations, common reasons:

• Task execution failed (e.g., LLM API call failed, network timeout)
• Estimated cost higher than actual cost (e.g., estimated 100 credits, actually used 80 credits, refund 20 credits)

All refund records have detailed descriptions, you can click to view the reason.

4. Data Security

4.1 Is my data safe?

OctoReport takes data security very seriously and employs multiple protection measures:

• Data Isolation: Each user's data is completely isolated, other users cannot access it
• API Key Encryption: All sensitive API Keys (Firecrawl, OpenAI, etc.) are encrypted and stored using AES-256-GCM
• Transmission Encryption: All data transmission uses HTTPS encryption
• Access Control: Strict user permission control, only administrators can access system configuration
• Regular Backups: Database is automatically backed up daily to prevent data loss

4.2 Will my content be used to train AI models?

No. OctoReport only sends content to third-party LLM services (OpenAI, Claude, etc.) for report generation and conversation, not for model training:

• OpenAI API does not use user data to train models by default (refer to OpenAI Data Usage Policy)
• Anthropic Claude API also has strict data privacy protection policies
• If using self-hosted LLM (future support), data will never leave your server

4.3 Can I export my data?

Yes, but currently requires manual export:

• Report Content: Go to the report details page and copy the Markdown content
• Library Content: Go to the library details page and copy the content list
• Ask Records: Go to the Ask session and copy the conversation content

One-click export functionality (JSON/CSV/Markdown formats) will be provided in the future.

4.4 Can data be recovered after deletion?

Depends on the deletion type:

• Soft Delete (e.g., deleting data sources, libraries, reports): Data is retained for 30 days, during which you can contact customer support for recovery
• Hard Delete (e.g., deleting account): Data is immediately physically deleted and cannot be recovered
• Expired Content: Content marked as expired during URL deduplication is retained for 90 days, then automatically deleted

5. Usage Limits

5.1 What are the usage limits?

To ensure system stability and fairness, the following limits are set:

If you need to increase the limit, please contact customer support.

5.2 What is the minimum execution interval?

Minimum execution interval for data sources and report templates:

• Data Sources: Minimum interval of 1 hour (to avoid frequent scraping and resource waste)
• Report Templates: Minimum interval of 1 hour (to avoid frequent report generation)

If you need more frequent execution, it is recommended to use the "Manual Trigger" feature.

6. Contact Support

6.1 How to contact customer support?

We provide multiple support channels:

6.2 What information should I provide when reporting an issue?

To resolve issues faster, please provide the following information:

• Issue Description: Detailed description of the issue and steps to reproduce
• Screenshots or Screen Recording: If it's a UI issue, provide screenshots or screen recording
• Error Logs: If there are error messages, provide complete error information
• Task ID: If it's a data source or report issue, provide the task ID from the task log
• Browser and Operating System: For example, "Chrome 120 / macOS 14.0"

6.3 How to submit feature suggestions?

We welcome user feature suggestions! Submission methods:

• Email: Send to [email protected], mark the subject as "Feature Suggestion"
• Online Feedback: Submit feedback on the "Settings" page, select type as "Feature Suggestion"
• GitHub Discussions: Start a discussion on GitHub

💡We regularly review user suggestions, and high-priority features will be implemented in the next version.

7. Technical Architecture FAQ

In-depth answers about OctoReport's core technology, covering architecture design, billing mechanisms, security measures, and more.

7.1 Why choose RSSHub hybrid architecture?

A single RSSHub instance has a single point of failure risk. We adopt a hybrid architecture of "embedded instance + external instance" with automatic switching by priority, ensuring collection success rate > 95%. Even if the embedded instance fails, it can seamlessly switch to external instances.

7.2 How do URL deduplication and expiration work with report generation?

When a duplicate URL is successfully collected, the old version is marked as "expired" and the new version is saved. Report generation by default only uses non-expired content, avoiding confusion from multiple versions of the same URL. Historical versions can still be traced but do not participate in reports.

See URL Deduplication Technical Documentation.

7.3 How to ensure billing consistency?

We use database transactions + row locking mechanism to ensure "check balance → charge → execute task" is an atomic operation. Tasks automatically rollback on failure, never resulting in "charged but task failed". All transaction records are fully traceable and auditable.

See Atomic Billing Technical Documentation.

7.4 How are task failures and retries presented?

Tasks are retried up to 3 times after failure, with status clearly marked as "RETRYING". After the 3rd failure, it is marked as "FAILED" and an email notification is sent. Each attempt is logged independently (attempt count, execution time, error reason) for easy troubleshooting.

7.5 Is private deployment supported?

Fully supported. OctoReport is an MIT License open-source project that can be self-hosted on any server. We provide complete deployment documentation (one-click Docker Compose deployment) and technical support (GitHub Issues).

7.6 How are security and log protection implemented?

• API Key Encryption: AES-256-GCM encrypted storage
• Multi-tenant Isolation: All queries filtered by userId by default
• Log Desensitization: Production logs automatically filter Cookie/Session/Token
• Access Control: ADMIN/USER roles, route-level middleware protection

7.7 What's the difference from regular "LLM aggregation websites"?

Regular aggregation websites only do "webpage collage", we do "engineering-trustworthy data processing":

• Data Deduplication & Version Management: Avoid duplicate/outdated content
• Atomic Billing & Compensation: Never lose money
• Task Queue & Retry: Stable and reliable
• Complete Logs & Observability: Traceable issues

7.8 Which LLM models are supported?

All OpenAI SDK compatible models, including:

• OpenAI: GPT-4o, GPT-4o-mini, o1
• Volcano Engine: Doubao series
• Anthropic: Claude 3.5 Sonnet
• Deepseek: Deepseek V3
• Custom: Any OpenAI API compatible model

7.9 How does the Credit system charge?

Billing by operation type:

• Search/Crawler: $0.01-0.05 per operation (depending on source type)
• LLM Cleaning: $0.02-0.1 per operation (depending on model)
• Report Generation: Billed by tokens (model pricing)

All fees are estimated in advance, tasks are blocked when balance is insufficient. See "3.2 How are credits consumed" above.

7.10 Can data be exported?

Absolutely. All data is stored in PostgreSQL database, supporting:

• Database Export: pg_dump complete backup
• API Export: Get JSON data via API
• Report Download: Markdown/HTML formats

7.11 How to contribute code or report issues?

Contributions welcome!

• GitHub Repository: github.com/Octo-o-o-o/OctoReport
• Issue Submission: github.com/Octo-o-o-o/OctoReport/issues
• Pull Request: Follow code standards, pass tests to merge

7.12 Is technical support available?

• Community Support: GitHub Issues (free)
• Email Support: [email protected] (cloud service users priority)
• Technical Documentation: Complete deployment guide, API reference, troubleshooting

8. More Resources

Have other questions? Send an email to [email protected], we will reply as soon as possible.