feat(config): add workspace.deferPublish to skip publishers on release

[BryanFRD]

@

What

Adds workspace.deferPublish (bool, default false). When true, ferrflow release skips the inline publisher phase; ferrflow publish still runs the publishers regardless.

Why

Publishers run at the end of ferrflow release, inside whatever job runs the release. When that job is minimal (no docker buildx / helm / built npm dist/), repos had to keep a separate config file for ferrflow publish so the release run wouldn't try to publish and fail. That two-file split is awkward. With deferPublish: true, a repo keeps one .ferrflow (publishers + the flag): release versions/tags, and a separate ferrflow publish job (with the toolchain) publishes.

Notes

• Default false → existing behaviour unchanged (publishers run on release, as Kit relies on).
• Flag is read from config by the binary — no action / reusable-workflow changes needed.
• Implemented as a guard in the release post-publish phase; ferrflow publish ignores the flag.

Verification

Unit tests (parse + default). Functional check in a temp repo: ferrflow --dry-run release with deferPublish: true shows no publisher output, while ferrflow --dry-run publish runs them. Full suite + clippy green; schema mirrored to FerrFlow-Cloud.

Closes #582
@

[github-actions]

Benchmark

Details

Benchmark suite	Current: 74c68d6	Previous: 0b5fe57	Ratio
version_files/toml_read	8130 ns/iter (± 59)	8489 ns/iter (± 791)	0.96
version_files/toml_write	183557 ns/iter (± 19643)	132036 ns/iter (± 10286)	1.39
version_files/json_read	5522 ns/iter (± 41)	5314 ns/iter (± 69)	1.04
version_files/json_write	137826 ns/iter (± 25747)	130851 ns/iter (± 8755)	1.05
version_files/xml_read	5164 ns/iter (± 46)	5140 ns/iter (± 16)	1.00
version_files/xml_write	129313 ns/iter (± 9980)	138826 ns/iter (± 18123)	0.93
version_files/gradle_read	5028 ns/iter (± 58)	4981 ns/iter (± 37)	1.01
version_files/gradle_write	129980 ns/iter (± 9352)	128857 ns/iter (± 12186)	1.01

This comment was automatically generated by workflow using github-action-benchmark.

[github-actions]

⚠️ Performance Alert ⚠️

Possible performance regression was detected for benchmark.
Benchmark result of this commit is worse than the previous benchmark result exceeding threshold 1.20.

Benchmark suite	Current: 74c68d6	Previous: 0b5fe57	Ratio
version_files/toml_write	183557 ns/iter (± 19643)	132036 ns/iter (± 10286)	1.39

This comment was automatically generated by workflow using github-action-benchmark.