planned
Public timelines across Mastodon instances through each instance’s API. Planned.
When Mastodon ships as a source, discatch will read public timelines across instances through each instance’s own API, free within its rate limits. Hashtag search works everywhere; full-text search depends on the instance. Instance scoping is honest: discatch checks the instances in its list, not the entire fediverse. Mastodon is planned.
Public posts on the instances discatch knows, with hashtag search everywhere and full-text search where an instance enables it.
Each instance’s own API, free within its rate limits, fanned out through the same instances table Discourse uses.
One syntax for every source: quotes, AND, OR, NOT, site:, author:, lang:.
query
#yourproduct OR "your product"One result, as a catch delivers it: source, author, the post, a prescore, thread context, and a suggested action your agent can act on.
Looking for a tool that monitors fediverse mentions of my open-source project. F5Bot does HN and Reddit, but nothing does Mastodon. Surely someone has built this?
Mastodon’s default search is hashtag-only. An instance administrator can enable full-text search by running an Elasticsearch or Meilisearch backend, but many do not. discatch uses what each instance offers: hashtag search where that is all there is, full-text where available. The result says which instance it came from, so you know the coverage.
discatch’s instances table is shared between Discourse and Mastodon. It ships seeded with well-known instances and grows as requests name new ones. A bare domain is enough to add one. Up to 20 instances per request, each priced in the quote. The table tracks each instance’s search capability so discatch knows whether to send a hashtag-only query or a full-text query.
The default Mastodon rate limit is 300 requests per 5 minutes per instance. discatch distributes requests across instances and respects each one’s limits. If an instance is rate-limited, the quote reflects it and that instance’s results arrive on the next interval. The quote says what is available right now.
Zero dedicated vendor tool pages exist for Mastodon keyword alerts or monitoring. A GitHub issue from 2018 asking for keyword-triggered content warnings ranks number one. Syften’s homepage appears for one phrasing but has no dedicated page. The rest of the results are platform docs, Apify actors, and an OSINT blog. Developer communities that moved to the fediverse have no listening tool today.
Mastodon is planned. No results arrive from Mastodon today. Some fediverse content may appear as web results if the index has picked it up, but this is not reliable. When Mastodon ships, discatch will check each instance’s public API on the interval you set and push new matches to your agent.
Mastodon is not one site. It is thousands of instances, each with its own API and its own search capabilities. discatch fans requests out to the instances in its list, seeded with well-known ones and extended when a request names a new one with site:. Many instances support only hashtag search. Full-text search depends on the instance administrator. This is an honest constraint that discatch states rather than hides.
Hashtag and product name
#yourproduct OR "your product"Watch one instance
site:mastodon.social "your product"Mastodon uses the same instances-table pattern as Discourse. discatch keeps a list of instances, seeded with well-known ones, and adds a new one the first time a request names it with site:. A request fans out to up to 20 instances, and the quote prices each. This is the same architecture, applied to a different federated system.
The difference is search capability. Every Discourse instance has full-text search through its JSON API. Many Mastodon instances only support hashtag search. Full-text search depends on whether the instance administrator has opted in. discatch uses what each instance offers and states this honestly on every result.
Open-source maintainers, Rust and Nix communities, privacy advocates, and parts of the Ruby and Elixir ecosystems are active on Mastodon. If your product serves these communities, their conversations about you happen on instances you may not know about. A catch across the instance list with your product name finds them.
No. Mastodon is planned. It is behind the same gate as the other planned sources.
A seeded list of well-known instances, extendable on request. Name one with site: in the query. Up to 20 instances per request.
No. Many Mastodon instances only support hashtag search. Full-text search depends on whether the instance administrator has enabled it. discatch uses what each instance offers.
The same units as any catch: one alert per firing plus one result per item delivered. Each instance in the request is priced in the quote.
Planned, yes. Like Discourse, a new instance is added the first time a request names it with site:. A bare domain is enough.
discatch reads each instance’s own API, fanned out across the instances table. It does not federate a query across the network; it checks each instance individually.
Not directly. Some Mastodon content may appear as web results if the index has picked it up, but this is not reliable.
None. Zero dedicated vendor tool pages exist for "monitor Mastodon for mentions" or "Mastodon keyword alerts." Syften’s homepage ranks for one phrasing. The results are platform docs, Apify actors, and a GitHub issue from 2018 asking for keyword-triggered content warnings.
No. Mastodon uses per-instance polling, not a firehose. The catch interval determines how often discatch checks each instance.
Threads is not planned. Threads caps keyword search at 500 queries per week, which makes it impractical as a monitoring source.
Bluesky has a free firehose that delivers every post in real time. Mastodon has per-instance APIs, each with its own rate limits and search capabilities. Bluesky is one network; Mastodon is thousands of instances. Both are planned; Bluesky will be cheaper and faster.
Last reviewed by Parsa Khazaeepoul, founder, dcouple. Facts from docs/data-sources.md and docs/coverage-matrix.md; status from docs/api.md; the search landscape from docs/seo/keyword-map.md.