Epic: rosetta cross-language consistency — per-language deviation tracking (46 languages) #2560

One tracking issue per language for the keyword-rosetta control corpus's measured cross-language bias (corpus built for #1096; gitgalaxy-side CI integration is #2557). Identical planted intent in all 46 languages — every deviation is measured engine bias, validated in the corpus's deviation ledger.

Current standing: 3 of 33 metrics pass the strict cross-language gate (no language beyond ±50% of median); on average 75% of languages sit within ±25% per metric (bias report, chart).

Strategy: structure first, then keywords, then risk

Each child issue is organized in fix order:

1. Structure detection — function/args/class detection, dependency links, imports. Where these are broken (ada and embedded_python report functions_found: 0; html's func_start is dead end-to-end per html func_start is unreachable end-to-end: the polyglot splitter consumes <script>/<style> tags before html's rules run #2549), every downstream number for that language is built on sand. Fix these first.
2. Signal/keyword rules — per-keyword over/under-counting (literal shielding prism/detector: string-literal shielding is selective by signal family (if/try/open count inside a literal, eval does not) #2535, dual-rule overlaps, return-in-branch kotlin: branch rule counts 'return' — inconsistent with java/swift/c siblings #2545, etc.).
3. Risk-exposure consequences — downstream of 1–2; expected to re-baseline substantially once the upstream layers fix. Don't tune risk formulas against today's biased inputs.

The improvement loop is cyclic and cheap: fix an engine issue → re-run the corpus gates + tools/bias_report.py → the ledger/report/chart regenerate → re-baseline expected_signals.json where intended (precedent: keyword-rosetta 0e8ebfa after #2548) → tick the affected languages here. Cross-cutting fixes (#2535 touches ~40 languages, #2546 most of them) will move many checkboxes at once — expect the list to collapse in clusters, not one-by-one.

Tier-2 caveat: the worst-looking rows (markdown 29🔴, css/html 14🔴, jcl 13🔴, yaml/yacc) are mostly zero-signal morphology (markdown genuinely has no io/state/telemetry), not extraction bugs — their fix is largely scoring-side (risk formulas treating absent morphology as incomparable rather than divergent), plus the real bugs marked in their issues (#2549 for html).

Cross-cutting engine issues feeding this epic

#2535 (selective literal shielding) · #2536 (hit_vector opacity) · #2537 (hyphen-identifier debt leak) · #2539 (js with) · #2540 (fortran USE case) · #2541 (php open tag) · #2542 (go import bypass) · #2543 (swift 1-char imports) · #2544 (swift open) · #2545 (return-in-branch family) · #2546 (×3 flux weighting) · #2547 (orphan conversion opacity) · #2549 (html func_start) — #2538 already fixed.

Per-language tracking (worst first)

• rosetta[markdown]: cross-language consistency — 29 red / 1 amber metrics, 3 ledgered shapes #2587 markdown — 29🔴 / 1🟡, 3 shapes — structure-level red (Tier-2 morphology)
• rosetta[html]: cross-language consistency — 14 red / 7 amber metrics, 3 ledgered shapes #2578 html — 14🔴 / 7🟡, 3 shapes — structure-level red (Tier-2 morphology)
• rosetta[css]: cross-language consistency — 14 red / 5 amber metrics, 2 ledgered shapes #2570 css — 14🔴 / 5🟡, 2 shapes — structure-level red (Tier-2 morphology)
• rosetta[jcl]: cross-language consistency — 13 red / 7 amber metrics, 2 ledgered shapes #2581 jcl — 13🔴 / 7🟡, 2 shapes — structure-level red (Tier-2 morphology)
• rosetta[yaml]: cross-language consistency — 9 red / 8 amber metrics, 2 ledgered shapes #2606 yaml — 9🔴 / 8🟡, 2 shapes — structure-level red (Tier-2 morphology)
• rosetta[yacc]: cross-language consistency — 9 red / 3 amber metrics, 3 ledgered shapes #2605 yacc — 9🔴 / 3🟡, 3 shapes — structure-level red (Tier-2 morphology)
• rosetta[m4]: cross-language consistency — 8 red / 6 amber metrics, 4 ledgered shapes #2585 m4 — 8🔴 / 6🟡, 4 shapes — structure-level red
• rosetta[sqlite]: cross-language consistency — 8 red / 2 amber metrics, 4 ledgered shapes #2601 sqlite — 8🔴 / 2🟡, 4 shapes — structure-level red
• rosetta[apex]: cross-language consistency — 7 red / 5 amber metrics, 6 ledgered shapes #2564 apex — 7🔴 / 5🟡, 6 shapes — structure-level red
• rosetta[haskell]: cross-language consistency — 6 red / 4 amber metrics, 4 ledgered shapes #2577 haskell — 6🔴 / 4🟡, 4 shapes
• rosetta[dockerfile]: cross-language consistency — 5 red / 11 amber metrics, 5 ledgered shapes #2572 dockerfile — 5🔴 / 11🟡, 5 shapes — structure-level red
• rosetta[embedded_python]: cross-language consistency — 5 red / 6 amber metrics, 4 ledgered shapes #2573 embedded_python — 5🔴 / 6🟡, 4 shapes — structure-level red
• rosetta[fortran]: cross-language consistency — 5 red / 4 amber metrics, 5 ledgered shapes #2574 fortran — 5🔴 / 4🟡, 5 shapes
• rosetta[agc_assembly]: cross-language consistency — 5 red / 3 amber metrics, 3 ledgered shapes #2563 agc_assembly — 5🔴 / 3🟡, 3 shapes — structure-level red
• rosetta[matlab]: cross-language consistency — 5 red / 3 amber metrics, 5 ledgered shapes #2588 matlab — 5🔴 / 3🟡, 5 shapes
• rosetta[python]: cross-language consistency — 5 red / 3 amber metrics, 5 ledgered shapes #2593 python — 5🔴 / 3🟡, 5 shapes
• rosetta[scheme]: cross-language consistency — 5 red / 3 amber metrics, 5 ledgered shapes #2598 scheme — 5🔴 / 3🟡, 5 shapes — structure-level red
• rosetta[solidity]: cross-language consistency — 5 red / 1 amber metrics, 4 ledgered shapes #2600 solidity — 5🔴 / 1🟡, 4 shapes (Tier-2 morphology)
• rosetta[swift]: cross-language consistency — 5 red / 0 amber metrics, 4 ledgered shapes #2602 swift — 5🔴 / 0🟡, 4 shapes — structure-level red
• rosetta[cobol]: cross-language consistency — 4 red / 8 amber metrics, 5 ledgered shapes #2567 cobol — 4🔴 / 8🟡, 5 shapes — structure-level red
• rosetta[perl]: cross-language consistency — 4 red / 6 amber metrics, 5 ledgered shapes #2590 perl — 4🔴 / 6🟡, 5 shapes
• rosetta[livecode]: cross-language consistency — 4 red / 5 amber metrics, 5 ledgered shapes #2583 livecode — 4🔴 / 5🟡, 5 shapes
• rosetta[powershell]: cross-language consistency — 4 red / 4 amber metrics, 6 ledgered shapes #2592 powershell — 4🔴 / 4🟡, 6 shapes
• rosetta[c]: cross-language consistency — 4 red / 3 amber metrics, 4 ledgered shapes #2566 c — 4🔴 / 3🟡, 4 shapes
• rosetta[lua]: cross-language consistency — 4 red / 2 amber metrics, 3 ledgered shapes #2584 lua — 4🔴 / 2🟡, 3 shapes
• rosetta[zig]: cross-language consistency — 4 red / 2 amber metrics, 5 ledgered shapes #2607 zig — 4🔴 / 2🟡, 5 shapes
• rosetta[makefile]: cross-language consistency — 3 red / 4 amber metrics, 4 ledgered shapes #2586 makefile — 3🔴 / 4🟡, 4 shapes — structure-level red
• rosetta[abap]: cross-language consistency — 3 red / 3 amber metrics, 2 ledgered shapes #2561 abap — 3🔴 / 3🟡, 2 shapes
• rosetta[ada]: cross-language consistency — 3 red / 2 amber metrics, 4 ledgered shapes #2562 ada — 3🔴 / 2🟡, 4 shapes — structure-level red
• rosetta[objective-c]: cross-language consistency — 3 red / 1 amber metrics, 6 ledgered shapes #2589 objective-c — 3🔴 / 1🟡, 6 shapes — structure-level red
• rosetta[go]: cross-language consistency — 3 red / 0 amber metrics, 5 ledgered shapes #2575 go — 3🔴 / 0🟡, 5 shapes
• rosetta[tcl]: cross-language consistency — 2 red / 3 amber metrics, 4 ledgered shapes #2603 tcl — 2🔴 / 3🟡, 4 shapes
• rosetta[assembly]: cross-language consistency — 2 red / 2 amber metrics, 5 ledgered shapes #2565 assembly — 2🔴 / 2🟡, 5 shapes
• rosetta[groovy]: cross-language consistency — 2 red / 2 amber metrics, 4 ledgered shapes #2576 groovy — 2🔴 / 2🟡, 4 shapes
• rosetta[kotlin]: cross-language consistency — 2 red / 2 amber metrics, 5 ledgered shapes #2582 kotlin — 2🔴 / 2🟡, 5 shapes
• rosetta[php]: cross-language consistency — 2 red / 2 amber metrics, 4 ledgered shapes #2591 php — 2🔴 / 2🟡, 4 shapes
• rosetta[scala]: cross-language consistency — 2 red / 2 amber metrics, 3 ledgered shapes #2596 scala — 2🔴 / 2🟡, 3 shapes
• rosetta[javascript]: cross-language consistency — 1 red / 5 amber metrics, 4 ledgered shapes #2580 javascript — 1🔴 / 5🟡, 4 shapes
• rosetta[shell]: cross-language consistency — 1 red / 5 amber metrics, 5 ledgered shapes #2599 shell — 1🔴 / 5🟡, 5 shapes
• rosetta[java]: cross-language consistency — 1 red / 2 amber metrics, 3 ledgered shapes #2579 java — 1🔴 / 2🟡, 3 shapes
• rosetta[dart]: cross-language consistency — 1 red / 1 amber metrics, 4 ledgered shapes #2571 dart — 1🔴 / 1🟡, 4 shapes
• rosetta[typescript]: cross-language consistency — 1 red / 0 amber metrics, 4 ledgered shapes #2604 typescript — 1🔴 / 0🟡, 4 shapes
• rosetta[cpp]: cross-language consistency — 0 red / 2 amber metrics, 3 ledgered shapes #2568 cpp — 0🔴 / 2🟡, 3 shapes
• rosetta[csharp]: cross-language consistency — 0 red / 2 amber metrics, 3 ledgered shapes #2569 csharp — 0🔴 / 2🟡, 3 shapes
• rosetta[ruby]: cross-language consistency — 0 red / 2 amber metrics, 3 ledgered shapes #2594 ruby — 0🔴 / 2🟡, 3 shapes
• rosetta[rust]: cross-language consistency — 0 red / 0 amber metrics, 2 ledgered shapes #2595 rust — 0🔴 / 0🟡, 2 shapes

A language's issue closes when it sits inside the ±25% band on every comparable metric, or every remaining deviation is ledgered as intended morphology.