Prepare next release

[github-actions]

This PR was opened by the Changesets release GitHub action. When you're ready to do a release, you can merge this and the packages will be published to npm automatically. If you're not ready to do a release yet, that's fine, whenever you add more changesets to main, this PR will be updated.

Releases

@germ-network/atprotoclient@0.8.0

Minor Changes

• #46 be79b49 Thanks @germ-mark! - Serve com.atproto.repo.createRecord from MockPDS. Same guards and body handling as putRecord, but the PDS mints the record key — a TID, so the key callers recover from uri parses as one. A caller-supplied rkey is refused with a 400 rather than honored: without the already-exists failure modeled too, honoring it would just be putRecord under another name. Production paths that create records (blocking, for one) were previously untestable against the mock: it 400'd on the way in.

  Fixes the record uri the mock reports. Reads returned at://did:web:example.com/NSID(rawValue: "app.bsky.graph.block")/<rkey> — a hardcoded authority, and collection interpolated as a struct rather than its rawValue — and putRecord returned the bare placeholder "example.com". Every uri is now built in one place from the repo's own DID, so create, put and read all name the same record the same way.

  Fixes the cid that createRecord and putRecord return: it was the literal "mock", which does not parse as a CID (no b prefix, not base32), so feeding it back to anything that takes one failed at the boundary. It is now Atproto.CID.mock().string, matching what reads already returned.

  Raises the GermConvenience floor to 0.3.0 and aligns the mock's errors with its cleaned-up handling. The mock's generic 400s said "Invalid Request" — with a space, matching no atproto error name — so parse fell through and every one reached the caller as an opaque .unrecognized(400 ). They are InvalidRequest now, which is in defaultErrors, so consumers get a typed .xrpcError they can match on. getRecord's catch moved off HTTPResponseError.unsuccessfulString, which 0.3.0 no longer throws, onto the type plus its code / bodyString accessors.

  Breaking: MockRepo.init takes the repo's did. Callers that construct a MockRepo directly need updating; MockPDS.host(did:bskyProfile:) is unchanged. Anything asserting on the old uri or cid values, or matching a mock 400 as .unrecognized, needs updating too.

  The GermConvenience floor is the one to watch downstream: SwiftPM's from: is upToNextMajor on 0.x, so this drags a consumer's whole graph onto 0.3.0, which is source-breaking for anything that mutates a BundledHTTPRequest. oauth4swift ≥ 0.6.0 carries its companion change; first-party mutation sites migrate to settingHeader(_:for:).