[Rafael-SOWNet]

abs(1/x^pi) at 0 was NaN, the claim that the limit does not exist. It is +oo from both sides, as 1/abs(x)^pi is.

The modulus took its argument's limit and nothing else. To the left of 0, x^pi is complex, so there was no reading of which sign 1/x^pi's divisor vanishes with. That side had no limit, and the two-sided limit was NaN. 2.6.0 gives the same results.

Now, where the argument's limit gives the modulus no value, the modulus of a quotient is read as the quotient of the moduli, and of a product as the product. Both hold for complex values. Read that way, the divisor vanishes from above, and the existing vanishing-divisor reading answers it. A modulus that already has a limit as written keeps it, and the tests pin three of those.

Input, at 0	2.6.0 and master	Now
abs(1/x^pi)	NaN; unevaluated from the left	+oo
abs(1/sqrt(x))	NaN	+oo
abs(2/x^(3/2)) from the left	unevaluated	+oo
abs(log(cos(x), -1)), which is π/|ln cos x|	NaN; unevaluated from the right	+oo

Measured

• limitcheck at DEPTH=3 on 21 seeds (14 to 51), master against this branch: 3 more rows compared, across 2 seeds. Those are the two NaNs above, abs(1/x^pi) at seed 51 and abs(log(cos(x), -1)) at seed 41. No other row changed.
• The new tests fail on master: the three infinite rows I ran before adding the fourth. The three "keeps what it had" rows pass on both.
• Suite: 18988 passed, 0 failed. AOT passed. BoundCheck, RootCheck, CasBench, PropCheck, SimpSweep, RuleCheck, CanonCheck and Confluence are clean.

Recorded in BREAKING-CHANGES.md.

🤖 Generated with Claude Code

[codecov]

Codecov Report

❌ Patch coverage is 75.00000% with 3 lines in your changes missing coverage. Please review.
✅ Project coverage is 86.77%. Comparing base (fdc586f) to head (4676d3d).

Files with missing lines	Patch %	Lines
...nctions/Continuous/Limits/Solvers/Limit.Classes.cs	75.00%	1 Missing and 2 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #2371      +/-   ##
==========================================
+ Coverage   86.76%   86.77%   +0.01%     
==========================================
  Files         312      312              
  Lines       68897    68908      +11     
  Branches    20078    20081       +3     
==========================================
+ Hits        59777    59798      +21     
+ Misses       4923     4912      -11     
- Partials     4197     4198       +1     

Flag	Coverage Δ
cpp	37.77% <ø> (ø)
csharp	87.32% <75.00%> (+0.01%)	⬆️
fsharp	32.97% <ø> (ø)
interactive	40.72% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

Files with missing lines	Coverage Δ
...nctions/Continuous/Limits/Solvers/Limit.Classes.cs	74.23% <75.00%> (+0.29%)	⬆️

... and 4 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[Rafael-SOWNet]

Merged without the three macOS jobs. They were cancelled before they started ("The job was not acquired by Runner of type hosted even after multiple attempts"), the same macOS arm64 outage as on #2368–#2370. In their place:

• I ran this PR's rows on a 512 KB thread, the tightest stack the macOS leg gives, along with x * ln(x). Each limit came out as the tests expect.
• The branch sits on current master, so there was nothing to trial-merge.
• Every Linux and Windows check is green.
• The local chain, suite and harnesses all passed.