State of issues/enhancements etc as of today by claude, spot checked. Will use this to draft up some loose roadmap-py thing to tackle tech debt and finalise the LS migration . I am going to go over it again tomorrow, because I am fairly certain there is another set of duplicates/already completed issues in there.

Can I ask what prompt you gave?

something like this:

⏺ Sure, here's the history of what we've done in this session:

  1. Initial request: Triage all open issues, PRs, bugs and feature requests for sarek. Close irrelevant/old/addressed stuff, mark won't-fix, close stale PRs, classify remaining as quick wins
   or not, pull bugs into a project board, merge related issues, come up with a plan including syntax compliance.
  2. Phase 1a: Reviewed 3 wontfix issues still open - you closed all 3.
  3. Phase 1b: Reviewed 21 stale moreinfo issues (no response >6 months) - you closed 19, kept #1020 (ASCAT stabilization) and #1555 (DragMap).
  4. Phase 1b addendum: Reviewed 5 more recent moreinfo issues - you closed #1133/#1134/#1575, kept #1741 (GZI index) and #2110 (join mismatch cluster).
  5. Phase 1c: Reviewed 5 old question/support issues - you closed 4, kept #1396.
  6. Phase 1d - Old enhancements (batch 1, #97-#879): You kept the majority, closed a few manually. Flagged #830 as your next project, #845 as easy fix, #848 for docs, #863 for CNV calling
  category, #877 as easy fix.
  7. Phase 1d - Old enhancements (batch 2, #880-#984): Tracked #908 and #923 in haplotypecaller improvements bucket. Kept most, closed some manually.
  8. Phase 1d - Old enhancements (batch 3, #989-#1076): Kept or closed manually.
  9. Phase 1d - Old enhancements (batch 4, #1152-#1346): Kept or closed manually.
  10. Phase 1d - Old enhancements (batch 5, #1354-#1576): Flagged #1392 (module output encapsulation) as important for later code updates. Kept or closed manually.
  11. Phase 1d - Old bugs (36 issues): Kept or closed manually.
  12. Phase 1d - Internal housekeeping (14 issues): Kept or closed manually.
  13. Phase 2 - Stale PRs (26 candidates): You kept most as somewhat valuable despite merge conflicts. Closed some manually.
  14. Phase 3 - Categorization: Pulled fresh issue list (238 remaining, down from 304). Categorized all issues into: Quick Win, Syntax Compliance, Optimization, Giant Genomes, tool groups
  (ASCAT, CNVkit, GATK/HC, VEP/snpEff, etc.), bug groups (Join Mismatch, Input/Samplesheet, CRAM/BAM, etc.), and other categories.
  15. Category refinements: You requested #1822 -> GATK/HC, #1141 -> Optimization, #1018 -> Giant Genomes, #868 -> joint genotyping/GATK, #852/#848 -> Documentation. Added new buckets: New
  Tool Requests, Preprocessing/Alignment, Platform/Infrastructure, QC Enhancements, Freebayes, MSISensor, MultiQC.
  16. Final cleanup: Re-pulled latest state (238 issues), verified all categorized, eliminated uncategorized bucket entirely. Opened final report in VS Code.

my initial prompt was along the lines of:

Sarek is an open source project. We recently have piled up a lot of technical and project managment debt. I need help triaging all open issues, PRs, bugs and feature requests. My basic idea is: close all irreleveant, super old, and/or adressed stuff. mark anything we won't fix as such and close. Close stale PRs that won't be merged. Then take the remaining issues, classify them as quickwins, or not, pull all the legitimate bugs into a project board. Ideally merge any issues that are related. Then come up with a plan on how to adress this, as well as required changes such as achieving compliance with the strict syntax and topics channels

and then use the above to fine tune as I went on

I was more wondering if we could do the same on other pipelines/modules

oh I think so. Let me cross post in team-maintainers