chore(internal): bump IR to 67.2.0 for SseStreamChunk.resumable

[patrickthornton]

Summary

• Bump packages/ir-sdk/fern/apis/ir-types-latest/VERSION from 67.1.0 → 67.2.0 and append a CHANGELOG entry. PR feat(parser): x-fern-streaming.resumable for SSE auto-reconnection #15795 added an optional resumable field to SseStreamChunk but did not bump the IR version, so no new @fern-fern/ir-sdk was published and the field stayed invisible to downstream consumers (e.g., the Go SDK follow-up that wants to read it).
• Documentation update for bumping IR versions.

Doc update (packages/ir-sdk/CLAUDE.md, "Adding New IR Features" → new step 4):

Every PR that modifies ir-types-latest/definition/ must bump the IR version, otherwise no new IR package is published and consumers can't pick up the change.

• Minor bump (67.1.0 → 67.2.0): additive, backward-compatible changes — new optional fields, additive type declarations. This is the common case.
• Patch bump: non-schema fixes (doc-string edits, internal refactors).
• Major bump: breaking changes — batched and done as a separate "Update IR to N+1" PR, not per-feature.

[claude]

Claude Code Review

This repository is configured for manual code reviews. Comment @claude review to trigger a review and subscribe this PR to future pushes, or @claude review once for a one-time review.

_{Tip: disable this comment in your organization's Code Review settings.}

[github-actions]

SDK Generation Benchmark Results

Comparing PR branch against median of 5 nightly run(s) on main (latest: 2026-05-08T04:59:46Z).

Full benchmark table (click to expand)

Generator	Spec	main (generator)	main (E2E)	PR (generator)	Delta
csharp-sdk	square	65s (n=5)	96s (n=5)	60s	-5s (-7.7%)
go-sdk	square	124s (n=5)	274s (n=5)	122s	-2s (-1.6%)
java-sdk	square	209s (n=5)	271s (n=4)	198s	-11s (-5.3%)
php-sdk	square	55s (n=5)	77s (n=5)	49s	-6s (-10.9%)
python-sdk	square	127s (n=5)	230s (n=5)	126s	-1s (-0.8%)
ruby-sdk-v2	square	77s (n=5)	111s (n=5)	81s	+4s (+5.2%)
rust-sdk	square	157s (n=5)	157s (n=5)	148s	-9s (-5.7%)
swift-sdk	square	51s (n=5)	736s (n=2)	49s	-2s (-3.9%)
ts-sdk	square	80s (n=5)	100s (n=5)	83s	+3s (+3.8%)

main (generator): generator-only time via --skip-scripts (includes Docker image build, container startup, IR parsing, and code generation — this is the same Docker-based flow customers use via fern generate). main (E2E): full customer-observable time including build/test scripts (nightly baseline, informational). Delta is computed against generator-only baseline.
⚠️ = generation exited with a non-zero exit code (timing may not reflect a successful run).
Baseline from nightly runs on main (latest: 2026-05-08T04:59:46Z). Trigger benchmark-baseline to refresh.
Last updated: 2026-05-08 16:30 UTC

[github-actions]

Docs Generation Benchmark Results

Comparing PR branch against median of 5 nightly run(s) on main (latest: 2026-05-08T04:59:46Z).

Fixture	main	PR	Delta
docs	337.2s (n=5)	305.3s (35 versions)	-31.9s (-9.5%)

Docs generation runs fern generate --docs --preview end-to-end against the benchmark fixture with 35 API versions (each version: markdown processing + OpenAPI-to-IR + FDR upload).
Delta is computed against the nightly baseline on main.
Baseline from nightly run(s) on main (latest: 2026-05-08T04:59:46Z). Trigger benchmark-baseline to refresh.
Last updated: 2026-05-08 16:30 UTC