АРЧ АП11 - Архитектура: вынести reply packaging из composeStage в отдельный owner

docs/ARCH/11 - architecture_turnaround/08 - current_status_audit_2026-04-17.md

@@ -25,15 +25,15 @@ This snapshot is based on:
 
 Latest graph rebuild:
 
-- `5228 nodes`
-- `11338 edges`
-- `133 communities`
+- `5251 nodes`
+- `11337 edges`
+- `136 communities`
 
 Most relevant current god nodes for turnaround `11`:
 
 1. `resolveAddressIntent()`  
-2. `ChannelRegistry`  
-3. `composeFactualReply()`  
+2. `composeFactualReplyBody()`  
+3. `ChannelRegistry`  
 4. `CanonicalStore`  
 5. `compactWhitespace()`
 
@@ -42,7 +42,7 @@ The relevant conclusion is not that every god node is part of turnaround `11`.
 The relevant conclusion is:
 
 - `resolveAddressIntent()` remains the main unresolved domain-intent concentration point;
-- `composeFactualReply()` remains the main unresolved answer-shaping concentration point;
+- `composeFactualReplyBody()` now carries the remaining answer-shaping concentration after packaging extraction;
 - `assistantService` still appears as a large coordinator-heavy community rather than a thin shell.
 
 ## What Is Already Real In Code
@@ -126,7 +126,7 @@ This is enough to build targeted semantic packs that are not single-domain toy s
 
 ## Honest Phase Status
 
-Estimated overall turnaround completion: `~87%`
+Estimated overall turnaround completion: `~88%`
 
 ### Phase 0. Shared Baseline
 
@@ -177,17 +177,19 @@ Remaining debt:
 
 ### Phase 4. Coverage / Evidence / Truth Gate Isolation
 
-Status: `84%`
+Status: `86%`
 
 Reason:
 
 - explicit truth and coverage/evidence contracts exist;
 - answer policy reads those contracts rather than rebuilding verdicts blindly from raw rows.
+- reply-packaging mechanics are now explicitly split into `address_runtime/replyPackaging.ts` instead of staying fully in `composeStage.ts`.
 
 Remaining debt:
 
-- `composeFactualReply()` is still a major concentration point;
-- humanized blocked/limited semantics are not yet fully separated from final packaging logic across all paths.
+- `composeFactualReplyBody()` is still a major concentration point;
+- humanized blocked/limited semantics are not yet fully separated from answer semantics across all paths;
+- `composeStage.ts` still remains too large even after packaging extraction.
 
 ### Phase 5. AssistantService Extraction
 
@@ -244,6 +246,7 @@ Compared with the pre-turnaround baseline, the system is now materially better i
 - meta questions and memory recap are no longer purely incidental side effects of route logic;
 - organization data-scope probing is no longer owned only by coordinator-local helper bodies;
 - debug payload assembly is now further isolated from top-level turn coordination;
+- reply formatting and reply-type classification now have an explicit owner outside `composeStage.ts`;
 - architecture regressions can now be localized to route, transition, truth gate, coverage/evidence, boundary, or meta/memory layers.
 
 ## What Still Remains The Main Architectural Debt
@@ -258,9 +261,9 @@ Intent resolution remains one of the most connected business nodes in the graph.
 
 This means capability and contour growth still concentrate pressure there.
 
-### 3. `composeFactualReply()` is still too central
+### 3. `composeFactualReplyBody()` is still too central
 
-Truth contracts are now explicit, but final answer-shaping still retains too much architecture weight.
+Truth contracts are now explicit, and reply packaging has started moving into its own owner, but final answer-shaping still retains too much architecture weight.
 
 This is the main remaining reason why user-facing humanization and limitation semantics are not completely isolated yet.
 
@@ -281,7 +284,7 @@ But not every business family has reached the same contract maturity.
 The next honest architecture slice should be:
 
 1. continue reducing `assistantService.ts` to a thinner coordinator;
-2. isolate answer-shaping semantics further away from `composeFactualReply()`;
+2. continue isolating answer semantics further away from `composeFactualReply()` now that reply packaging has its own owner seam;
 3. keep extending AGENT packs with mixed business + meta + interruption patterns instead of single-family smoke tests;
 4. keep using scenario acceptance as the main sign-off rather than unit-test green status alone.