Frequently Asked Questions
Find clear guidance on configuring feeds, managing integrations, setting up AI agents, duplicate detection, and custom filters on PusheRSS.
PusheRSS monitors your configured RSS and Atom feeds in the background. When new items are published, it automatically formats, filters, optionally transforms them with AI models, and dispatches them to your connected integration destinations (such as Discord webhooks, custom HTTP webhooks, SMTP emails, or aggregated RSS feeds).
Navigate to the Feeds tab in your dashboard and click Add New Feed. Enter your RSS/Atom feed URL, an optional display name, a badge color, and select which integrations should receive updates from this feed.
Go to the Integrations tab and click Add New Integration. Select your desired plugin type (Discord, Custom Webhook, Email SMTP, or Combined RSS Feed), configure the destination settings, attach any desired keyword or regex filters, and select connected feeds.
You can click the Duplicate button on any integration card in the Integrations list or on the Integration edit page. This instantly clones the configuration, badge color, attached filters, and connected feed destinations so you can easily create similar setups.
When 'Check for duplicate' is enabled on a feed, PusheRSS checks incoming items against previous dispatch history for that feed. If an item matches an existing title, link, or GUID/ID, it is prevented from re-dispatching to your integrations and logged in history with status 'Duplicate'.
Filters allow you to keep or drop items based on matching rules. Create filters under the Filters tab by choosing a type (Filter In to keep matching items, or Filter Out to drop matching items) and a rule (Contains for keyword matching or Matches for regular expressions). You can attach any number of defined filters directly when creating or editing an Integration.
AI Agents allow you to transform raw RSS feed items using Large Language Models before notifications are delivered. Instead of receiving raw article excerpts, an AI Agent can synthesize multiple feed items into an editorial news digest, provide analytical deep dives explaining the background and implications, or translate content into your preferred language.
PusheRSS currently supports two dedicated AI Agent processing modes:
- Brief Mode: Acts as an expert news editor. It clusters related stories, filters low-value noise, and generates a structured, scannable summary digest.
- Explain Mode: Acts as a senior news analyst. It unpacks historical background, breaks down complex topics, and explains key implications and impact.
You can also set a specific output Language (e.g. English, French, Spanish, German, Japanese, etc.) or leave it blank to match the original content language. Prompt formatting is automatically optimized for your destination plugin.
1. Navigate to the AI Agent tab in the dashboard navbar and click Add New Agent Configuration.
2. Select your AI provider (Google Gemini, OpenAI, Claude, Mistral AI, DeepSeek, Perplexity, or xAI Grok).
3. Enter a name, select your default mode and language, enter your provider API key, and choose your preferred model.
4. Save the agent (PusheRSS automatically verifies your API key and model with a ping test).
5. Open any Integration (Discord, Email, Webhook), scroll to Agentic processing, and select your configured Agent.
You can obtain API keys directly from each provider's developer console:
- Google Gemini: Sign in at Google AI Studio, click Create API key, and select or create a project.
- DeepSeek: Sign in at DeepSeek Platform, navigate to API Keys, and click Create new API key.
- Anthropic (Claude): Sign in at Anthropic Console, go to Settings > API Keys, and click Create Key.
- Mistral AI: Sign in at Mistral AI Console, navigate to API Keys, and click Create new key.
- OpenAI (ChatGPT): Sign in at OpenAI Platform and click Create new secret key.
- Perplexity: Sign in at Perplexity AI Settings, ensure API credits are active, and click Generate Key.
- xAI (Grok): Sign in at xAI Console, navigate to API Keys, and click Create API Key.
Important Notice on AI Token Consumption & Non-Responsibility
1. Direct Third-Party Billing: When an AI Agent is enabled, PusheRSS directly contacts your chosen AI provider using your personal API key. All input tokens (feed item titles, descriptions, links, system prompts) and generated output tokens are billed directly to your account by the AI provider at their current pricing rates.
2. Token Consumption Factors: Token consumption varies depending on how frequently your monitored RSS feeds update, how much text is contained in each article, and the specific model you choose (e.g. flagship models vs cost-effective mini/flash models).
3. PusheRSS Non-Responsibility Disclaimer: PusheRSS assumes no liability or responsibility for any third-party AI costs, token consumption, billing overages, API quotas, rate limiting, provider service outages, or the accuracy of content generated by external AI models.
4. Recommended Spending Limits: We strongly advise setting monthly spending limits, credit caps, or billing alerts directly in your AI provider's account dashboard to prevent unexpected charges.
Feeds are polled every minute in the background. Only valid feed items published within the last 7 days that haven't been previously dispatched are processed.
You can view recent notification history (last 7 days) on your main Dashboard overview as well as on individual feed and integration detail pages. Click the Resend button next to any history entry to re-trigger that dispatch.
The main Dashboard overview features a Dispatch Activity graph at the bottom showing daily dispatches over the last 7 days grouped by integration, with color indicators and detailed tooltips for each date.
Some integrations support 'Scheduled Dispatch' mode, which allows you to define a cron schedule for notifications instead of sending them immediately as new items arrive. You can configure this setting within the Integration edit form for supported plugins.
We respect your privacy. PusheRSS uses a minimal set of necessary cookies for essential functionality, such as maintaining your login session. We may also use non-essential analytics cookies, which you can manage at any time using the cookie settings.
If you have any questions, feedback, or need further assistance, please feel free to reach out to us at pusherss@senorihl.link.