scan_mentionsFind every mention. 55+ platforms. One call.

If your agent's job touches "what is the world saying about X right now", it needs a single call that fans out across the open social, news, and community web — and brings back deduplicated mentions with timestamps, links, and snippets. scan_mentions does that. Reddit, X, Bluesky, Mastodon, Hacker News, Substack, podcasts, reviews, news, GitHub, Stack Overflow. Four credits. About three seconds.

When to call this tool

This is the workhorse of the catalog. Any agent doing competitive research, brand monitoring, journalism, lead-discovery, or post-launch listening should call this first. It's the cheapest tool that returns the broadest signal. Pair it with score_intent when the consumer is a sales motion, with get_dossier when the consumer is a human briefing, and with run_geo_audit when the consumer is a marketing motion.

The tool deduplicates aggressively across mirrors. A Reddit post that gets re-shared to X and quoted in a Substack will return as one canonical mention with three source links rather than three separate rows. The lookback window defaults to 7 days; agents can extend up to 90 with the lookback_days argument. There is no "real-time push" mode; agents that need that should poll on a schedule.

Input schema

{
  "query": "string  // brand, person, term, or boolean expression",
  "lookback_days": "integer  // 1-90, default 7",
  "platforms": ["string  // optional, restrict to subset"],
  "languages": ["string  // optional, ISO codes"],
  "limit": "integer  // 1-200, default 50"
}

Output schema

{
  "query": "string",
  "window": { "start", "end" },
  "mentions": [
    {
      "platform", "author", "published_at",
      "snippet", "link", "engagement": { "likes", "comments" }
    }
  ],
  "summary": { "total", "by_platform", "top_authors" },
  "credits_used": 4
}

Example invocations

1. Claude Desktop

Find every mention of "Cursor" on Reddit and Hacker News
in the last 14 days. Group by sentiment.

2. ChatGPT custom GPT

For "what is the internet saying about X" prompts, call
scan_mentions with the brand and a 7-day lookback.

3. Cursor MCP

Use scan_mentions to pull GitHub issues mentioning
"OpenTelemetry" in the last 30 days. Save as JSON.

4. n8n

{ "jsonrpc": "2.0", "method": "tools/call",
  "params": { "name": "scan_mentions",
    "arguments": { "query": "Anthropic OR Claude", "lookback_days": 3 } } }

5. Raw curl

curl -X POST https://www.mentionfox.com/mcp \
  -H "Authorization: Bearer $FOXAPIS_KEY" \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"scan_mentions","arguments":{"query":"Resend","lookback_days":7}}}'

Sample output

{
  "query": "Resend",
  "summary": { "total": 142, "by_platform": { "x": 61, "reddit": 38, "hn": 14 } },
  "mentions": [
    { "platform": "hn", "author": "swyx", "snippet": "We swapped Postmark for Resend last week..." }
  ],
  "credits_used": 4
}

Credit cost & rate limits

Flat 4 credits regardless of result count, up to the 200-row hard cap. 60 calls per minute per key. The query string supports OR / AND / NOT and quoted phrases.

Error codes & recovery