Put the new crate in CI, before it is the one that breaks

`diaswarm-peer` is a fifth crate and CI knew about four. That is exactly the
gap this workflow was added to close this morning — `diaswarm-net` and
`diaswarm-android` had never run in CI, and one of them broke within hours.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Author: alex-poor

.github/workflows/crates.yml

@@ -92,6 +92,10 @@ jobs:
       # mDNS between peers inside the runner. If this one turns out to be flaky
       # on a hosted runner it must be made reliable or made explicit — not
       # quietly dropped back out, which is the state it is being rescued from.
+      - name: diaswarm-peer
+        working-directory: crates/diaswarm-peer
+        run: cargo test --no-fail-fast
+
       - name: diaswarm-net
         working-directory: crates/diaswarm-net
         run: cargo test --no-fail-fast