# Second CMP screening working document

This version focuses on MP29 (HS+20), MP45 (HSsym), and MP43 (HS−20), using the same saved 500-iteration results and format as the first paper. Fixed-catch artifacts are reused unchanged from the first document. The general all-CMP scorecard remains available for context; the selected scorecard contains the three requested CMPs plus the fixed-catch comparison.

From the jmMSE26 source checkout, after generating the current advice probabilities and the first paper:

```sh
export JMMSE_CMP_PAPER_OUT=output/cmp-working-paper-hs-2026
export JMMSE_CMP_FOCUS=tun29,tun45,tun43
export JMMSE_SCORECARD_FOCUS='HS+20 (MP29)|HSsym (MP45)|HS-20 (MP43)|Fixed catch (1,525 kt)'
export JMMSE_REUSE_FIXED_CATCH=true
Rscript R/build_cmp_screening_paper.R
Rscript R/extend_cmp_paper_robustness_fixedcatch.R
Rscript R/build_hs_screening_tables.R
python3 R/configure_cmp_scorecard_oms.py
quarto render output/cmp-working-paper-hs-2026/cmp-screening-working-paper.qmd
quarto render output/cmp-working-paper-hs-2026/scorecard-all.qmd
quarto render output/cmp-working-paper-hs-2026/scorecard-shortlist.qmd
```

The editable Quarto narrative belongs to this second comparison and must remain with its own figures. The source checkout also requires the sibling saved-run directories documented by the first paper. This workflow reads those runs and does not refit or retune models.

## Recruitment-cycle and near-term update

The focused checkpoint set is now OM11_2 and OM11_3. `R/build_cmp_nearterm.R` generates the new 2026–2035 trade-off/Kobe plots for both CMP sets and checks the reference Kobe values against the established calculation. Run it after the focused robustness build, then render both papers and regenerate their scorecards. The demonstration selectors use reference, recruitment crash and recruitment cycle.

```sh
Rscript R/build_cmp_nearterm.R
```