Track a brand, product or topic across developer communities in one run: Stack Exchange questions, Lemmy posts and Mastodon hashtag activity, merged into one schema with text, direct link, author, community and engagement numbers. Official/open APIs, no key needed.
Initial release. One keyword, three developer communities, one schema: Stack Exchange (any site in the network, 100 results per page with deep paging and server-side date filters), Lemmy (public API, paged) and Mastodon (hashtag timelines, paged by max_id). Every row carries the text, a direct link, the author, the community and that source's engagement numbers, plus a per-source report in the key-value store.
Honest source note: Mastodon's full-text search requires an account token, so this Actor reads hashtag timelines — free-text mentions without a hashtag are not covered, and the README says so.
A source that fails is recorded in the per-source report and the run continues with the others; the run only fails if every source fails.
The result budget is shared between the selected sources (with unused budget passing on), after a test run came back 400 Stack Exchange rows and nothing from Lemmy or Mastodon — selecting three communities has to return three communities.
The default time window is all time with newest-first sorting, chosen after measuring that a 30-day window can be genuinely near-empty on Stack Overflow (2 questions matching "supabase" in 30 days) — recent mentions still come first, but a quiet month no longer looks like a broken Actor.
