fix(proto): bump to 0.2.2 to unblock the stalled release

[SebastianThiebaud]

Why

Release run 26369479219 failed at cargo publish --verify for tsoracle-server 0.2.3:

error[E0432]: unresolved imports `tsoracle_proto::v1::LEADER_HINT_TRAILER_KEY`,
              `tsoracle_proto::v1::LeaderHintLookup`
error[E0425]: cannot find function `decode_leader_hint` in module `tsoracle_proto::v1`
error: failed to verify package tarball

Root cause: #295 lifted LEADER_HINT_TRAILER_KEY, LeaderHintLookup, and decode_leader_hint into tsoracle_proto::v1 under a refactor: prefix. Crates version independently and release_commits = "^(feat|fix|perf)", so refactor: did not trigger a release — proto stayed at 0.2.1, and the published 0.2.1 (from release #233) predates and lacks those symbols. tsoracle-server 0.2.3 already uses them, so the verify build — which strips the path dep and resolves tsoracle-proto ^0.2.1 from the registry — got the stale crate and failed to compile.

What

• tsoracle-proto → 0.2.2 (patch: additive, non-breaking). All consumers pin ^0.2.1, so they resolve to 0.2.2 with no cascade. Once this lands on main, the release job publishes proto 0.2.2 ahead of the stuck dependents (server 0.2.3, client 0.2.2, …), which then verify against a proto that has the symbols.
• CONTRIBUTING.md — document the trap (library-crate API additions must be feat:/fix:, never refactor:, because the API is visible to dependent crates even when the v1 wire contract is unchanged), correct the stale "versions move in lockstep" claim (independent since chore(release): version crates independently to fix release-plz resolution #223), and spell out the release_commits prefix filter.

Notes

• The release job publishes on a version-vs-registry diff, so the bump republishes regardless of this PR's squash prefix.
• The release job does not regenerate changelogs, so tsoracle-proto's CHANGELOG.md will have a gap at 0.2.2 — left as-is rather than hand-editing it.
• Follow-up worth considering: enable semver_check = true so the Release PR flags "dependent uses a symbol absent from the published dep."

[coveralls]

Coverage Report for CI Build 26370056716

Coverage remained the same at 94.913%

Details

• Coverage remained the same as the base build.
• Patch coverage: No coverable lines changed in this PR.
• No coverage regressions found.

Uncovered Changes

No uncovered changes found.

Coverage Regressions

No coverage regressions found.

---

Coverage Stats

Relevant Lines:	11873
Covered Lines:	11269
Line Coverage:	94.91%
Coverage Strength:	453420.47 hits per line

---

💛 - Coveralls