docs: document 0.1.4 analysis foundation

CHANGELOG.md

@@ -4,6 +4,16 @@
 
 Development opened after stable v0.1.3 release.
 
+- Implemented the loopback-only Web Analyse V1 workspace and shared bounded
+  Core read model. The 7/30/90-day and all-history filters, factual activity,
+  active-Program progress, compatible exercise metrics, explicit persisted
+  MAX results, primary BODY ZONE exposures, and fourteen body measurements are
+  calculated by the C17 backend. Dashboard summaries consume the same model
+  while preserving all existing tile identities and layout behavior. Deep
+  links, safe query fallbacks, null/empty states, French/English presentation,
+  keyboard focus, and responsive charts are covered without estimating 1RM,
+  readiness, fatigue, recovery, body fat, or performance.
+
 ## 0.1.3 — 2026-09-20
 
 - Decoupled the Android full-generation request from the strict legacy V3

README.md

@@ -350,7 +350,8 @@ As of 2026-09-18:
   `WEB_DASHBOARD_V1=PASS/FROZEN` and `TRAINLOG_WEB_SESSIONS_V1=PASS/FROZEN`;
   `/programmes` is an operational active-Program calendar, while
   Sessions → Programmes remains the administration surface;
-  `TRAINLOG_WEB_EXERCISES_V1=PASS`; Analyse remains a placeholder;
+  `TRAINLOG_WEB_EXERCISES_V1=PASS`; `/analyse` now provides the 0.1.4 factual
+  Dashboard/Analyse read-model foundation pending visual review;
 - `TRAINLOG_PROGRAMS_PRESENTATION_ANDROID_DELETE_V1=PASS`: the coordinated
   private desktop/Android deployment, real Firefox presentation, correlated
   `programs-v1` deletion ACK, restart/replay, and non-resurrection checks pass;

docs/architecture.md

@@ -61,6 +61,14 @@ long-term history. Core owns data, persistence, business logic, calculations,
 read models, command services, and versioned contracts. An interface must not
 query tables directly to reconstruct its own business rules.
 
+The Web Dashboard and Analyse workspace share `GET /api/v1/analysis`, backed by
+one C17 read model. Period filtering, active-Program progress, compatible
+exercise metrics, explicit MAX results, primary BODY ZONE exposures, and body
+measurement first/last/delta calculations occur in Core. React validates the
+bounded response and presents it; it does not load raw history or issue SQL.
+Absent values remain null, and the model never estimates physiological or
+performance scores.
+
 Android local SQLite is a capture store, not a synchronization format.
 
 ## 2. Components

docs/current_state.md

@@ -141,7 +141,7 @@ parallel implementation of its rules.
 | Private full-generation rollout | `TRAINLOG_SYNC_FINALIZATION_AND_ROLLOUT_V1=PASS` (direct MTP, one authorized phone) |
 | Local Web | `TRAINLOG_WEB_V1=CONTRACT_FROZEN / IMPLEMENTATION_STARTED` |
 | Web Exercises V1 | `TRAINLOG_WEB_EXERCISES_V1=PASS` |
-| Current operational cursor | `TRAINLOG_WEB_ANALYSIS_V1` |
+| Current operational cursor | `TRAINLOG_WEB_ANALYSIS_V1` (implemented, visual review pending) |
 | Complete synchronization gap contract | Frozen dependency contract; operational USB/Drive slices required by v0.1.2 are delivered |
 | Isolated synchronization test environment | `TRAINLOG_SYNC_TEST_ENV_V1=PASS/FROZEN` |
 | Synchronization characterization | `TRAINLOG_SYNC_CHARACTERIZATION_V1=PASS/FROZEN` |
@@ -291,7 +291,24 @@ explicit proposal derivation and generation-backed Android delivery. The
 top-level `/programmes` route is an operational daily active-Program calendar;
 Sessions → Programmes remains the technical administration/import, list,
 detail, archive, and delete surface. `/exercices` is the operational desktop
-catalogue administration surface; Analyse remains an explicit placeholder.
+catalogue administration surface. `/analyse` is implemented with Overview,
+Exercises, Distribution, and Measurements sections over one bounded Core read
+model. Its global 7/30/90-day/all filter is applied in the backend; Dashboard
+uses the same 30-day calculations for activity, active-Program progress,
+exercise progression, measurements, and BODY ZONE links.
+
+The Analyse API validates period, exercise identity, measurement metric and
+page, rejects unknown or duplicate parameters, and bounds exercises and chart
+points. Missing facts remain null. Exercise output respects persisted
+tracking/load modes, explicit MAX rows are never estimated, and duration,
+distance, speed, external load, and external volume appear only when their
+source observations support them. BODY ZONES are labelled as recorded primary
+zone exposures/associated sets rather than physiological percentages. All
+fourteen stored body metrics expose period-first, period-last, and an absolute
+delta only with at least two points. The route and Dashboard deep links have
+safe invalid-query fallbacks, French/English labels, keyboard controls, and
+responsive no-overflow layouts. Visual inspection remains pending before any
+freeze marker.
 
 Exercises uses bounded Core-owned list/detail reads, canonical name
 normalization, BODY ZONES taxonomy, profiled creation/update commands and
@@ -587,9 +604,8 @@ tests, Android debug assembly, and 211 Android tests (207 passed, four optional
 historical-fixture skips). Strict desktop builds use Clang 22.1.8; the compared
 pre-existing GCC 16.2.1 Web warning gate remains documented rather than green.
 
-- `TRAINLOG_WEB_EXERCISES_V1=PASS`. The next planned work is
+- `TRAINLOG_WEB_EXERCISES_V1=PASS`. `TRAINLOG_WEB_ANALYSIS_V1` is implemented
-  `TRAINLOG_ANDROID_UI_REDESIGN_V1`; it remains unimplemented. Analyse remains
+  on the 0.1.4 branch and awaits the requested Dashboard/Analyse visual review.
-  a placeholder and is not a v0.1.3 priority.
 - `APP_SHELL_V1` retains a recorded legacy visual/accessibility review gate; its
   observations are inputs to, not decisions for, the new 0.1.3 Android contract.
 - Session Generator V1 is hidden while V2 planning semantics are developed.

docs/roadmap.md

@@ -43,9 +43,9 @@ tile rendering exist. The Dashboard's bounded progression and BODY ZONES
 visualizations and Sessions V1 are implemented. The top-level `/programmes`
 route is an operational daily active-Program calendar, while
 Sessions → Programmes remains the technical administration/import, list,
-detail, archive, and delete surface. Web Exercises V1 is implemented; Analyse
+detail, archive, and delete surface. Web Exercises V1 is implemented. The
-remains a placeholder outside this cycle unless a later explicit decision
+0.1.4 `TRAINLOG_WEB_ANALYSIS_V1` branch now owns the factual Dashboard/Analyse
-changes that order.
+read-model foundation and stops before merge for visual inspection.
 The corrective prepared-item projection, bounded phase-owned MTP outbox,
 Sessions presentation correction, local date preference and durable preparation
 withdrawal are implemented. The withdrawal uses the separate versioned
@@ -54,12 +54,12 @@ mobile snapshot or `TRAINLOG_FORMAT_V1` contracts.
 
 ## Current cursor
 
-The v0.1.3 development cursor proceeds from the closed stable v0.1.2 baseline
+The v0.1.4 cursor proceeds from stable v0.1.3 and its validated Web, Program
-and its validated Web, Program execution, USB/Drive generation, ACK, and
+execution, USB/Drive generation, ACK, and causal-deletion contracts without
-causal-deletion contracts without reopening those completed gates.
+reopening those completed gates.
 
 ```text
-CURRENT_OPERATIONAL_CURSOR=TRAINLOG_ANDROID_UI_REDESIGN_V1_CONTRACT
+CURRENT_OPERATIONAL_CURSOR=TRAINLOG_WEB_ANALYSIS_V1
 ```
 
 The focused `TRAINLOG_SYNC_CAUSAL_DELETE_V1_CLOSEOUT` is complete before this

docs/tests.md

@@ -1,5 +1,14 @@
 # Tests and validation
 
+`web_analysis` covers empty history, every 7/30/90/all period selector,
+loaded and unloaded repetitions, continuous duration/distance/speed, explicit
+persisted MAX, multiple primary BODY ZONES, partial measurements, one- and
+multiple-point measurement semantics, and active-Program projection. The HTTP
+suite exercises the real bounded endpoint plus invalid and duplicate query
+parameters. Vitest validates the typed parser, null preservation, route and
+section changes, period/exercise/metric selection, invalid deep-link fallback,
+critical French/English labels, and Dashboard layout regressions.
+
 `test_sync_drive_transport` proves Drive artifact-before-manifest and
 manifest-before-reference publication, rejects unsafe/AI namespaces, and keeps
 a partial generation invisible. Orchestrator coverage proves that Drive mirror