Chat with mentions
Turn your current feed filters into a grounded AI chat over the matching posts, with clear credit rules and practical workflows.
Chat with mentions packages the mentions that match your current filters into a prompt that opens in your chosen AI assistant (ChatGPT, Claude, and others). The model can reason over that slice of your listening data instead of you scrolling and copying posts by hand.
Credit Cost
You are charged once per generated session, when SnitchFeed creates the API-backed deeplink and data access for that chat—not per message inside the external AI tool.
Each successful generation costs 100 credits. That applies every time you:
- Open a provider (for example ChatGPT or Claude) from the dialog, or
- Use copy prompt, which also creates the same deeplink session behind the scenes.
Generating a second link (another provider, or again after changing filters or the prompt text) is a new charge.
Ongoing conversation in the AI app does not bill additional SnitchFeed credits by itself.
How to chat with mentions
Set your filters
On the Feed (or any view whose URL carries your filters—platform, relevance, tags, date range, listeners, keywords, and so on), choose the slice you care about. Only mentions that match this filter set are included in the chat context.
Open Chat with Mentions
Click Chat with Mentions (the sparkle-style control in the feed toolbar).
Choose a prompt
Under Choose Prompt:
- Pick a template to load a structured prompt, or
- Customize your prompt in the text area (optional). If you customize, keep the {API_ROUTE} placeholder so the prompt still points at SnitchFeed’s data endpoint for your filtered results.
Pick a model (or copy the prompt)
Under Pick a Model, choose an assistant. SnitchFeed opens it in a new tab with the prompt applied. Alternatively, use copy prompt if you prefer to paste the prompt elsewhere.
Narrow filters first—high relevance, specific tags, or a tight date window—when you want a focused briefing rather than a very large export context.
Ways to use it
Use a narrow filter set first so the model works from a coherent slice of mentions, then ask for summaries, themes, or next actions. Below are common patterns teams use.
GTM and sales
Focus on high intent or tagged buying signals, then ask the assistant to summarize objections, competitor mentions, and concrete asks. Turn the output into talk tracks, outbound angles, or a weekly pulse for your reps without rereading the raw feed.
Competitive intelligence
Filter to competitor brands, products, or keywords, then ask for positioning shifts, feature comparisons, pricing chatter, and risks people call out in the wild. Good for battlecards and exec briefings when you need narrative, not just a spreadsheet.
Product and UX
Restrict to mentions tagged or filtered to your product area—bugs, feature requests, confusion, or integrations—then ask for recurring themes, severity signals, and language customers actually use. Helps prioritize the roadmap and align internal vocabulary with the market.
Marketing and content
Mine questions, phrases, and objections from a relevant audience slice so landing copy, FAQs, campaigns, and SEO briefs match how people talk when they are not on your site. Especially useful after launches or messaging tests when volume spikes.
Comms and crisis monitoring
Tighten date range and listeners around an incident or narrative, then ask for a factual timeline, who is driving the conversation, and gaps in your response before you publish a statement or FAQ.
Reusable setups
Combine this with Saved views: open a saved view, refine filters if needed, then run Chat with Mentions on that exact audience so every run starts from the same definition of “what matters.”