boardreadyops (oaslananka) errors

13 recurring problems engineers report in boardreadyops (oaslananka), ranked by impact.

#32 by errors trackedECADPLM

Recurring boardreadyops (oaslananka) errors, by impact

  1. 1A live review displayed eight blocking high-severity findings, but the manually published release_runs rows saved decision=pass, because the upload endpoint only counted 'error' as blocking. The fix refuses a pass whenever high or error findings exist, even if the caller asks for pass.ECAD · Agent failure · 1 report74
  2. 2A board review page now links to an exact source-file compare, while BOM and geometry differences stay marked unavailable unless recorded separately, so a source diff is not passed off as a BOM or mechanical diff.PLM · Agent failure · 1 report66
  3. 3A PCB readiness review tool guessed each rule's domain from its name prefix, and checked against its own rule catalog, 11 of 49 rules were in the wrong category in finding rows and whole-run rollups.ECAD · Agent failure · 1 report66
  4. 4A real run on a board design shows 195 recorded findings, 25 per page. On page one, old BOM warnings about KiCad's virtual power symbols compete with findings on populated parts, while the findings tied to a marker on the PCB should be looked at first. The fix only reorders within a page and keeps every finding and waiver.ECAD · Agent failure · 1 report66
  5. 5A run review showed an empty KiCad preview because the head revision had no artwork even though the baseline had real snapshots.ECAD · Agent failure · 1 report66
  6. 6BoardReadyOps published live demos of a GitHub App that blocks a PR on Edge.Cuts problems, duplicate references and BOM risk. Adjacent product worth watching.ECAD · Agent failure · 1 report66
  7. 7BoardReadyOps review publish ignored the --project option, so in a multi-board workspace a review could include snapshots and findings from unrelated KiCad projects. A regression now proves only the chosen project contributes.ECAD · Agent failure · 1 report66
  8. 8Hosted KiCad reviews show Hardware surface diff details are not available whenever BOM or changed-file data was not persisted. The stopgap compares SHA-256 digests of the published SVGs and says plainly that this is not geometry, BOM or electrical evidence.ECAD · Agent failure · 1 report66
  9. 9In the zaptrace PCB and schematic viewer, a schematic-sourced MPN finding has no recorded coordinate, so clicking it could leave the user on the copper layer. The fix only navigates when the finding's .kicad_sch filename matches exactly one published sheet, and refuses to guess otherwise.ECAD · Agent failure · 1 report66
  10. 10On a hosted KiCad review, Visual Diff showed an all-black canvas with no legend; a user cannot tell if that means no visible pixel changes, a broken visual, or a passing release. The fix says plainly that black is not electrical or fab approval.ECAD · Agent failure · 1 report66
  11. 11On a real hosted run, KiCad CLI was not installed on the runner, and the DRC and ERC findings told the user to commit a source correction. The fix separates 'the check never ran' from 'the board has a defect' on the finding card.ECAD · Agent failure · 1 report66
  12. 12On a real KiCad fixture, a BOM resistor had no PCB footprint assigned at all, and the check printed 'PCB footprint .' The fix states plainly that the comparison cannot happen because the PCB side is missing.ECAD · Agent failure · 1 report66
  13. 13The PCB review tool's visual design view only ever loaded page 1 of findings, 25 per page, with no way to page on, so later findings could not be selected next to board artwork.ECAD · Agent failure · 1 report58

How it fails

  1. Agent failure13

Sourced from public community reports. Costs and delays on individual entries are modeled estimates, not measurements.