Five deploy repos must drop snapshot-generate-cmd before/with #343, and RAINIX_SHA must be bumped after

[thedavidmeister]

PR #343 (closing #341) removes the snapshot-generate-cmd input from rainix-tag-release entirely. Every caller passing it must drop that line, or its release workflow fails with an unexpected-input error the moment #343 merges.

The five callers, and the ordering hazard

Today's rainix default for the input is the freezing form, forge script ./script/Build.sol --sig "cutRelease()" && forge fmt.

repo	passes	dropping the line today
rain.factory.deploy	the freezing form	no-op (matches default)
rain.metadata.deploy	the freezing form	no-op
rain.deploy	the freezing form	no-op
rain.extrospection.deploy	the freezing form	no-op
rain.math.float.deploy	forge script ./script/Build.sol && forge fmt (NON-freezing)	changes behaviour — it would fall back to the freezing default

So the first four can drop the line now, safely, ahead of #343. rain.math.float.deploy must not drop it until #343 has merged, after which the workflow runs the non-freezing Build.sol run() itself and the line is redundant rather than the thing selecting that form.

Also required after #343 merges

RAINIX_SHA must be bumped, or the new byte-comparison and monotonicity checks stay inert: the workflow runs release-guard from the pinned SHA, whose guard predates them. The #341 crash is fixed by the merge; the added guarantees are not live until the pin moves.

Done when

• rain.factory.deploy drops the line
• rain.metadata.deploy drops the line
• rain.deploy drops the line
• rain.extrospection.deploy drops the line
• rain.math.float.deploy drops the line (after fix(tag-release): determinism check that never removes the frozen record #343)
• RAINIX_SHA bumped past fix(tag-release): determinism check that never removes the frozen record #343

🤖 Generated with Claude Code