tools: restore the line a background sweep left mutated (Refs #2161)

[gHashTag]

if d > 26 on master should be if d >= 26.

A full tri gates mutate sweep was running in the background when I staged with git add -A, and the mutant it was holding at that instant went into the commit, into the pull request, through 30 green checks, and onto master.

Behaviour is unaffected — this is the mutant proved equivalent over 525,918 points two ticks ago, and the line's own # mutant-equivalent: comment says exactly that.

That is what makes it worth fixing rather than shrugging at: the code stopped matching the comment directly above it and every test stayed green, which is the precise condition under which a wrong line survives indefinitely.

The guard protects the run, not the operator

mutate's dirty-tree refusal answers "is the tree clean before I start?". Nobody was asking "is a sweep running while I stage?", and git add -A cannot tell a mutant from an edit.

Three defences, cheapest first:

1. Stage paths, not -A, while a sweep runs.
2. Check target/.tri-mutating — it exists for exactly this and was sitting right there.
3. Read the summary git prints: the commit said 20 insertions, 1 deletion and I had written only insertions.

Third self-inflicted git loss this session (§119 has the other two) and the only one that reached master. Same shape each time: a command whose blast radius is the whole tree, run while my attention was on one file.

The detector was not a test — it was mutate refusing to start on a dirty tree the next time I ran it, one tick late.

§122.

Refs #2161

[github-actions]

📓 NotebookLM Notebook linked to this PR

• Notebook ID: b83263109fb055dc
• URL: https://notebooklm.google.com/notebook/b83263109fb055dc

This notebook contains session context, decisions, and artifacts for this work.

[github-actions]

PR Dashboard

Generated at: 2026-08-24 04:43:51 UTC

• tools: restore the line a background sweep left mutated (Refs #2161) #2666 tools: restore the line a background sweep left mutated (Refs #2161) | Created: 2026-08-24 | Updated: 2026-08-24 | 17 failing / 22 total
• Unbreak FPGA CI and correct two unbacked hardware claims #2081 Unbreak FPGA CI and correct two unbacked hardware claims | Created: 2026-08-11 | Updated: 2026-08-20 | 5 failing / 27 total

Summary

Status	Count
Total Open PRs	2
PRs with Failing Checks	1
PRs with All Checks Green	1
READY	0
FAILING	1
PENDING	0

Seal Status

• ⚠️ STALE -- sha256(compiler.rs)=e6333575ec10 != manifest seal=87e5cbd3ad94.
  The committed NMSE numbers were certified against an older compiler.rs.
  Run scripts/reseal-check.sh locally for the two-step reseal command (advisory; not a merge gate).