No user-submitted related servers found.
Scored across 2 tools
The two tools occupy distinct lifecycle stages: collect starts an async run and returns an id, while fetch_corpus retrieves the resulting artifact. Boundaries are clear, though fetch_corpus bundles retrieval, validation, status reporting, and authorization flagging into one call, which slightly blurs whether it is a poller or a fetch operation.
collect is a bare verb while fetch_corpus uses a verb_noun form, so the pair does not follow a single predictable pattern. Both names are readable and domain-relevant, but 'collect' vs 'fetch' are near-synonyms used for different stages, which adds mild confusion.
Two tools is thin for what is effectively an async job lifecycle (start, status, retrieve, authorize). The design folds status checking into fetch_corpus to compensate, but the surface still feels borderline minimal.
The start-and-retrieve loop is covered, but the emphasized approval model (corpus is 'NOT approved research', buildAuthorized is false for every corpus) has no tool to approve, verify, or re-check approval status. Listing or cancelling runs is also absent, leaving agents at a dead end when they need to act on the authorization signal.