f2a721cb1fb37acedc1b8954d30715389e02b8e6
Regenerating the production files for the V2.1 eFuse rework surfaced two steps that neither the skill nor CPL_ROTATIONS.md documented, both caught only by diffing against the previously-shipped CPL: - --exclude-dnp does not drop 'in_bom no' parts, so J4 and TP1-TP5 landed in the raw export although the shipped CPL excluded them. Filter to designators that have a BOM line; the two sets should end up equal. - kicad-cli emits signed rotations while the correction step's % 360 only normalises corrected parts. Mixing the two made the diff show 34 phantom rotation changes (270.0 vs -90.0 is the same angle), burying the 2 real ones. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Description
No description provided
Languages
Python
84.9%
C
12.4%
Shell
1.1%
Dockerfile
0.7%
Makefile
0.4%
Other
0.5%