Data updated 2026-10-23 10:06 UTC+8 MCP · 7 tools 中文
Sandbox demo Change anything: this copy of the demo data is yours alone, invisible to others and cleared when idle. Alert delivery and user management are off; words not in the data can be searched on the web live (3 per visitor).
· Self-host it for your brand →

Data sources · Collect

Where we listen, and how often

Switch sources on or off, collect now, schedule collection; every run is logged. Platforms we cannot see are listed below - no pretending to cover the whole web.

01Collect now

6 search terms × 6 enabled sources

ClaudeAnthropicModel Context ProtocolMCP serverOpenAIChatGPT

Collection runs in the background, usually 1–3 minutes; refresh the run log below for results. Last run: just now

02Auto-collect

The built-in schedule runs inside echowatch serve; cron with echowatch collect keeps working - both use the same path.

03Data sources

16 sources

Measured
SourceTypeStatusLast storedTotalLast runEnabled
bluesky code Fresh just now 130 +22 / 24
gdelt code — — — —
hn code Fresh just now 134 +8 / 16
reddit code Fresh just now 43 failed
rss code — — — —
web code — — — —
arxiv config · rss — — — —
gdelt-news config · json_api Fresh just now 89 failed
github-issues config · json_api — — — —
google-news config · rss Fresh just now 314 +66 / 77
google-news-zh config · rss — — — —
lemmy config · json_api Fresh just now 114 +8 / 13
mastodon config · json_api — — — —
npr-text config · html_list — — — —
stackoverflow config · json_api — — — —
techcrunch config · rss — — — —

Add a source with one TOML file in configs/ (RSS / JSON API / HTML list) - see configs/README.md.

04Run log

Latest 30

TimeSourceTermNewResult
2026-10-23 09:59lemmyChatGPT 8 / 13 OK
2026-10-23 09:59google-newsChatGPT 66 / 77 OK
2026-10-23 09:59gdelt-newsChatGPT 0 / 0 HTTPError: HTTP Error 429: Too Many Requests
2026-10-23 09:59redditChatGPT 0 / 0 HTTPError: HTTP Error 429: Too Many Requests
2026-10-23 09:59hnChatGPT 8 / 16 OK
2026-10-23 09:59blueskyChatGPT 22 / 24 OK
2026-10-23 09:59lemmyOpenAI 25 / 33 OK
2026-10-23 09:59google-newsOpenAI 80 / 87 OK
2026-10-23 09:59gdelt-newsOpenAI 0 / 0 HTTPError: HTTP Error 429: Too Many Requests
2026-10-23 09:59redditOpenAI 0 / 0 HTTPError: HTTP Error 429: Too Many Requests
2026-10-23 09:59hnOpenAI 41 / 49 OK
2026-10-23 09:59blueskyOpenAI 19 / 25 OK
2026-10-23 09:59lemmyMCP server 1 / 2 OK
2026-10-23 09:59google-newsMCP server 14 / 22 OK
2026-10-23 09:59gdelt-newsMCP server 14 / 18 OK
2026-10-23 09:59redditMCP server 0 / 0 HTTPError: HTTP Error 429: Too Many Requests
2026-10-23 09:59hnMCP server 14 / 15 OK
2026-10-23 09:59blueskyMCP server 19 / 25 OK
2026-10-23 09:59lemmyModel Context Protocol 0 / 0 OK
2026-10-23 09:59google-newsModel Context Protocol 20 / 20 OK
2026-10-23 09:59gdelt-newsModel Context Protocol 0 / 0 HTTPError: HTTP Error 429: Too Many Requests
2026-10-23 09:59redditModel Context Protocol 21 / 22 OK
2026-10-23 09:59hnModel Context Protocol 0 / 0 OK
2026-10-23 09:59blueskyModel Context Protocol 25 / 25 OK
2026-10-23 09:59lemmyAnthropic 30 / 35 OK
2026-10-23 09:59google-newsAnthropic 45 / 75 OK
2026-10-23 09:59gdelt-newsAnthropic 75 / 75 OK
2026-10-23 09:58redditAnthropic 0 / 0 HTTPError: HTTP Error 429: Too Many Requests
2026-10-23 09:58hnAnthropic 22 / 28 OK
2026-10-23 09:58blueskyAnthropic 20 / 24 OK

05Import CSV

Import CSV exports from Brandwatch, Meltwater, Talkwalker, Brand24 or a spreadsheet: title, link, text, author, date and source columns are recognized by name. Re-importing adds nothing twice.

06Blind spots

What we cannot see

Pending
PlatformWhy
X / Twitterpaid API only (pay-per-read since 2026-02); no free search
Instagram · Facebookno public search API; Meta Graph API covers only your own pages
TikTokResearch API restricted to approved academic use
LinkedInno public post search API
YouTubeneeds a Google API key (10,000 quota units/day); not bundled
WeChat · Weibo · Douyinno open search API; FTS5 tokenizer is not CJK-aware yet
Broadcast TV · radio · printlicensed data only
Private groups · DMs · paywalled contentnot public; never collected

Verdicts cover enabled sources only, never the platforms above. For those, import their official CSV exports.