Bump version to 1.27.3

[ChrisRackauckas]

🚧 UNREVIEWED — awaiting review by @ChrisRackauckas. Filed by an AI agent running as @ChrisRackauckas; Chris has not seen this.
Harness: Claude Code 2.1.301 · Model: claude-opus-5-5
Conversation: https://claude.ai/code/session_01VZjwf7Ux9YtTzwHGHMqnsq

Bumps version in Project.toml from 1.27.2 to 1.27.3 for a patch release. Unreleased since v1.27.2 (last version in the General registry):

• Restrict tuple nodeselect methods to integer indices #174 Restrict tuple nodeselect methods to integer indices (fixes the CartesianIndex regression introduced by Make tuple-node getindex type-stable; replace @test_broken @inferred with @inferred #172)
• Make tuple-node getindex type-stable; replace @test_broken @inferred with @inferred #172 Make getindex type-stable for heterogeneous tuple nodes (bug fix)
• Allow JET 0.12 in the QA environment for Julia 1.13 #171 Allow JET 0.12 in the QA environment (test-only)
• Add AirspeedVelocity benchmarking CI #170 Add AirspeedVelocity benchmarking CI (CI-only)

Patch, not minor: the user-facing changes are a type-stability fix and the fix for its regression; no new or changed public API.

Verification: branch rebased onto master after #174 merged; git log v1.27.2..HEAD lists exactly these four PRs plus the bump.

Risk assessment

• Risk: low
• Blast radius: version field only; the release ships the already-merged commits listed above.
• Evidence: one-line diff. The independent reviewer ran on Julia 1.13.2: nodeselect(([10,20],[1.5,2.5,3.5]), CartesianIndex(2), 1) returns 1.5 at v1.27.2, throws MethodError after Make tuple-node getindex type-stable; replace @test_broken @inferred with @inferred #172, returns 1.5 at this head; test/tuple_nodes.jl passes.
• Independent review: codex-cli 0.162.1 + gpt-6.1-sol rated it low, high confidence, verdict MERGE: the diff since v1.27.2 is exactly Add AirspeedVelocity benchmarking CI #170/Allow JET 0.12 in the QA environment for Julia 1.13 #171/Make tuple-node getindex type-stable; replace @test_broken @inferred with @inferred #172/Restrict tuple nodeselect methods to integer indices #174 plus the bump, the CartesianIndex regression is fixed, nothing breaking (local codex session 01a14316-17cb-7201-9b1e-d9aaf9e9431d). The earlier round's CHANGES verdict (CartesianIndex regression) is resolved by Restrict tuple nodeselect methods to integer indices #174.
• Merge: auto-merge candidate on green CI

---

🤖 Posted by an AI agent — harness: Claude Code 2.1.301 · model: claude-opus-5-5
Conversation: https://claude.ai/code/session_01VZjwf7Ux9YtTzwHGHMqnsq

🤖 Generated with Claude Code

https://claude.ai/code/session_01VZjwf7Ux9YtTzwHGHMqnsq