sourcecode 5.8.22__tar.gz → 5.8.24__tar.gz

{sourcecode-5.8.22 → sourcecode-5.8.24}/CHANGELOG.md

@@ -1,5 +1,83 @@
 # Changelog
 
+## [5.8.24] — 2026-08-22
+
+*The thirteenth audit-pass queue is closed entire: all six rows as separate
+commits, the regression first, with the full suite green at 12 329 tests. Three
+of the reported mechanisms are wider than reported and are corrected against
+measurement; one defect was found by the battery rather than by either.*
+
+### Fixed — the regression the pass led with
+
+- **`cache warm` warms every root view (`AUD-600-R01`)** — the warm built the
+  compact view only, so the first `--agent` call after an explicit warm paid a
+  full analysis: measured on `mall` (524 Java files) `--agent` 7.59 s → 0.34 s
+  (×22) and the raw contract view 3.20 s → 0.33 s (×9.7), against `--compact`
+  flat at 0.34 s. The field measured ×49.2 at ~3 000 files. Each view turns a
+  different set of analysers on and therefore has its own L1 core;
+  `cache_model.ROOT_VIEWS` is now the one authority the warm iterates and
+  `cache model` publishes a row per view. `--views` warms a subset, and a view
+  left cold is named rather than discovered by its first call.
+
+### Fixed — the new pack surface
+
+- **The gate composes `verify-edit` (`AUD-600-B01`)** — a pack named *Change
+  Safety Gate* did not compose the change-safety command, so a working tree
+  that had just gained `GET /internal/dump-config` returning `System.getenv(key)`
+  answered `NO_CHANGES` / exit 0. It now answers `UNVERIFIED` / exit 2 with the
+  route and its `access_policy`. The general half is the rule: every pack
+  publishes a `non_coverage` block, in its payload and in `pack list`, and the
+  battery asserts it over the axis and rule catalogues.
+- **`--compact` bounds the assessment pack (`AUD-600-B02`)** — the flag was
+  forwarded to the children and nothing else. It now substitutes each
+  component's own decision summary for its body: `mall` 608 K → 42 K estimated
+  tokens, `thingsboard` 8.5 K, both under the ceiling. A refusal may not deny a
+  variant the command declares, so `--format markdown` is derived as the
+  bounded rendering it is.
+- **The report keeps the figures the pack carries (`AUD-600-B03`)** — three
+  cells read `unknown` while the figure sat in a component payload the same
+  pack had written. One catalogue, `MARKDOWN_FIGURES`, read with the same
+  reader the bounded projection uses; every remaining `unknown` names what it
+  looked for.
+- **A write refusal is decided from `argv` (`AUD-600-D01`)** — `--output-dir`
+  inside a read-only tree cost 47.9 s before refusing. The destination is now
+  decided before any component runs and the run is served without its artefacts,
+  the shape `verify --init --no-write` established. Nothing is created inside
+  the analysed tree and every refusal is still announced with its path and type.
+
+### Added
+
+- **`ask schema packs-v1` (`AUD-600-D02`)** — what each pack composes, what it
+  does not, its verdict vocabulary, its exit codes and the aggregation rule that
+  produced the verdict, generated from the pack declarations. The rule was
+  inferable from one run and published nowhere.
+- **Per-component timing in the pack manifest** — `duration_ms` per component
+  and a `timing` rollup naming the slowest one, so a 47 s gate can be attributed
+  from outside. A clock is never an answer: `timing` is excluded from
+  `content_id`, or two runs of an unchanged tree would disagree.
+- **The fourth exit code for packs** — when every unverified component was cut
+  short by its budget the pack leaves with `75` and names them in
+  `components_cut_short`. A mixed run keeps the stronger `2` and the verdict
+  word never moves.
+
+## [5.8.23] — 2026-08-22
+
+*The external corpus follow-up is closed: pack output is durable by default,
+the gate verdict matrix is covered, and the 37-repository read-only baseline is
+recorded without overstating the seven scale-limited cases.*
+
+### Fixed
+
+- **Pack aggregate persistence (`AUD-599-C01`)** — `--output-dir` now writes
+  the aggregate as `pack.json`, `pack.yaml` or `pack.md`; an explicit
+  `--output FILE` remains authoritative.
+- **Gate verdict coverage (`AUD-599-T01`)** — controlled tests distinguish
+  `PASS`, `BLOCK` and `UNVERIFIED` without conflating their exit codes.
+- **Corpus audit follow-up (`AUD-599-P01`, `AUD-599-Q01`)** — the baseline and
+  before/after Git snapshots for all 37 repositories are documented. The seven
+  repeated timeouts are classified as a scale limitation, not claimed as a
+  regression without a comparable phase-level baseline.
+
 ## [5.8.22] — 2026-08-22
 
 *The eleventh audit-pass queue is closed entire: all fourteen open rows shipped

{sourcecode-5.8.22 → sourcecode-5.8.24}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: sourcecode
-Version: 5.8.22
+Version: 5.8.24
 Summary: Persistent structural context and ultra-fast repeated analysis for AI coding agents
 Project-URL: Homepage, https://github.com/HarounDominique/sourcecode
 Project-URL: Repository, https://github.com/HarounDominique/sourcecode
@@ -58,7 +58,7 @@ Description-Content-Type: text/markdown
 > question. All analysis is local and deterministic; see [Privacy](#configuration--privacy)
 > before inspecting a repository whose telemetry policy you need to control.
 
-![Version](https://img.shields.io/badge/version-5.8.22-blue)
+![Version](https://img.shields.io/badge/version-5.8.24-blue)
 ![Python](https://img.shields.io/badge/python-3.9%2B-green)
 
 Copyright (c) 2026 Dominique Haroun. All rights reserved. ASK Engine 5.7.0 and
@@ -86,12 +86,40 @@ workflow-shaped packs for users who want one consumable result:
 ```bash
 ask pack assessment . --profile prod --output-dir .ask/packs/assessment
 ask pack gate . --since origin/main --output-dir .ask/packs/gate
+ask pack assessment . --format markdown -o assessment.md
 ```
 
+For deployment-aware posture evidence, pass the artefacts that describe the
+environment being assessed:
+
+```bash
+ask pack assessment . --deployment k8s/deployment-prod.yaml \
+  --output-dir .ask/packs/assessment
+```
+
+The gate also renders native CI log annotations without changing its JSON
+contract:
+
+```bash
+ask pack gate . --since origin/main --format github-actions
+ask pack gate . --since origin/main --format gitlab
+```
+
+`PASS`, `BLOCK` and `UNVERIFIED` are data decisions. The process exit code is
+the CI signal; `--allow-unverified` can change only that exit code, never the
+reported verdict. Every pack carries a manifest with component invocations,
+source evidence and content fingerprints.
+
 The [Product Route](docs/PRODUCT-ROUTE.md) is normative: ASK focuses on
 deterministic evidence for Java/Spring modernization and safe change. Packs
 compose existing commands; they do not hide or replace them.
 
+For agent integrations, local MCP exposes the same pack evidence through
+`get_assessment_pack` and `get_change_safety_gate`. MCP defaults to a compact
+projection that preserves the decision and manifest; use the CLI with
+`--output-dir` to persist complete component artifacts. See
+[MCP pack evidence](docs/architecture/DESIGN-pack-mcp-api.md).
+
 **The cache is not a performance optimization. It is what makes ASK Engine usable as infrastructure rather than a one-off tool.**
 
 ---
@@ -148,7 +176,7 @@ set `ASK_MAX_RSS_MB=<megabytes>`; if the process exceeds it, ASK refuses the
 answer with `MEMORY_TOO_LARGE`, including the observed peak and limit. Leave it
 unset, or set it to `0`, to disable the ceiling explicitly.
 
-**Current performance status.** Release 5.8.22 retains the shared Canonical IR
+**Current performance status.** Release 5.8.24 retains the shared Canonical IR
 path used by `validation` and now passes that same CIR into labelled `data-exposure`
 runs; the no-label path remains parse-free. The one-model-per-declared-label path is
 also retained. These are implementation properties, not duration promises:
@@ -182,7 +210,7 @@ brew tap haroundominique/sourcecode && brew install sourcecode
 # pip / pipx
 pipx install sourcecode        # or: pip install sourcecode
 
-ask version                    # ask 5.8.22 — and, on a build that has aged,
+ask version                    # ask 5.8.24 — and, on a build that has aged,
                                # how many releases have probably shipped since
 ```
 

{sourcecode-5.8.22 → sourcecode-5.8.24}/README.md

@@ -10,7 +10,7 @@
 > question. All analysis is local and deterministic; see [Privacy](#configuration--privacy)
 > before inspecting a repository whose telemetry policy you need to control.
 
-![Version](https://img.shields.io/badge/version-5.8.22-blue)
+![Version](https://img.shields.io/badge/version-5.8.24-blue)
 ![Python](https://img.shields.io/badge/python-3.9%2B-green)
 
 Copyright (c) 2026 Dominique Haroun. All rights reserved. ASK Engine 5.7.0 and
@@ -38,12 +38,40 @@ workflow-shaped packs for users who want one consumable result:
 ```bash
 ask pack assessment . --profile prod --output-dir .ask/packs/assessment
 ask pack gate . --since origin/main --output-dir .ask/packs/gate
+ask pack assessment . --format markdown -o assessment.md
 ```
 
+For deployment-aware posture evidence, pass the artefacts that describe the
+environment being assessed:
+
+```bash
+ask pack assessment . --deployment k8s/deployment-prod.yaml \
+  --output-dir .ask/packs/assessment
+```
+
+The gate also renders native CI log annotations without changing its JSON
+contract:
+
+```bash
+ask pack gate . --since origin/main --format github-actions
+ask pack gate . --since origin/main --format gitlab
+```
+
+`PASS`, `BLOCK` and `UNVERIFIED` are data decisions. The process exit code is
+the CI signal; `--allow-unverified` can change only that exit code, never the
+reported verdict. Every pack carries a manifest with component invocations,
+source evidence and content fingerprints.
+
 The [Product Route](docs/PRODUCT-ROUTE.md) is normative: ASK focuses on
 deterministic evidence for Java/Spring modernization and safe change. Packs
 compose existing commands; they do not hide or replace them.
 
+For agent integrations, local MCP exposes the same pack evidence through
+`get_assessment_pack` and `get_change_safety_gate`. MCP defaults to a compact
+projection that preserves the decision and manifest; use the CLI with
+`--output-dir` to persist complete component artifacts. See
+[MCP pack evidence](docs/architecture/DESIGN-pack-mcp-api.md).
+
 **The cache is not a performance optimization. It is what makes ASK Engine usable as infrastructure rather than a one-off tool.**
 
 ---
@@ -100,7 +128,7 @@ set `ASK_MAX_RSS_MB=<megabytes>`; if the process exceeds it, ASK refuses the
 answer with `MEMORY_TOO_LARGE`, including the observed peak and limit. Leave it
 unset, or set it to `0`, to disable the ceiling explicitly.
 
-**Current performance status.** Release 5.8.22 retains the shared Canonical IR
+**Current performance status.** Release 5.8.24 retains the shared Canonical IR
 path used by `validation` and now passes that same CIR into labelled `data-exposure`
 runs; the no-label path remains parse-free. The one-model-per-declared-label path is
 also retained. These are implementation properties, not duration promises:
@@ -134,7 +162,7 @@ brew tap haroundominique/sourcecode && brew install sourcecode
 # pip / pipx
 pipx install sourcecode        # or: pip install sourcecode
 
-ask version                    # ask 5.8.22 — and, on a build that has aged,
+ask version                    # ask 5.8.24 — and, on a build that has aged,
                                # how many releases have probably shipped since
 ```
 

sourcecode-5.8.24/docs/AUDIT-2026-08-22-5.8.22.md

@@ -0,0 +1,96 @@
+# ASK Engine 5.8.22: auditoría metodológica y de regresión
+
+**Fecha:** 2026-08-22  
+**Artefacto auditado:** release `5.8.22`, `build_commit`
+`583c07c405358d3e76dd20968960f5ab928608fd`  
+**Sujeto principal:** MSAS / `banyan-v2`, HEAD `3dde0376`, 3.342 ficheros Java  
+**Registro relacionado:** [`docs/DEFECT-LEDGER.md`](DEFECT-LEDGER.md), entrada
+`AUD-598-X01` y filas derivadas
+
+## Alcance y método
+
+Se consolidan dos rondas independientes sobre la misma release:
+
+- Ronda A: MSAS, 57 invocaciones, 52 cronometradas, 3 `pr-impact` y 5 ayudas,
+  `ASK_READONLY=1`, presupuesto explícito y 0 escrituras.
+- Ronda B: banco de 16 repositorios Java/Spring, unas 95 invocaciones, Windows
+  11 Pro, pipx y Git Bash.
+
+Ambas rondas comprobaron estado Git antes y después. No se observaron
+tracebacks, regresiones funcionales ni escrituras no autorizadas.
+
+## Resultado ejecutivo
+
+| Ronda | Resultado | Veredicto del auditor |
+|---|---:|---|
+| A | `8.58/10` | Comprar y actualizar |
+| B | `8.8/10` | Adoptar sin reservas |
+
+La mejora es válida para fiabilidad y rendimiento general, pero no autoriza a
+cerrar automáticamente defectos del ledger. La propia auditoría demuestra que
+el artefacto empaquetado puede no contener los cierres que el árbol de trabajo
+declara.
+
+## Confirmado cerrado
+
+- `ask --help`: la regresión de reconstrucción del árbol Click queda corregida.
+- `selftest`: 12 filas, 0 errores y 0 bytes de stderr en la ronda de campo.
+- Determinismo de `posture`: `population_total=1908` y
+  `unconditional=1903` en cinco ejecuciones.
+- Las regresiones de 5.8.21 sobre `compare`, `endpoints --servlets` e `impact`
+  no reaparecen.
+- No-regresión funcional en el barrido de cierres anteriores.
+
+Estos cierres permanecen en el ledger histórico con su commit. No se duplican
+como nuevas filas.
+
+## Hallazgos abiertos
+
+El estado siguiente es el que debe usarse para planificación y release:
+
+| ID | Estado | Hallazgo |
+|---|---|---|
+| `AUD-598-R01` / `ASK-PERF-002` | `open` | `ask . --compact` sube a 214,6 s frente a 154,0 s en 5.8.21, con payload byte-identical; es una regresión condicionada a esa vista/escala. |
+| `AUD-598-X01` | `open` | El ledger del producto va por delante del wheel: cinco cierres declarados no se observan en el artefacto `583c07c`. |
+| `AUD-598-B01` / `ASK-META-001` | `open` | La auditoría de wheel no observa `_meta`, `content_id` ni memoria en 9/9 payloads inspeccionados. Debe resolverse dentro de un wheel construido, no sólo en checkout. |
+| `AUD-598-B02` / `ASK-UX-002` | `open` | `posture --diff` ofrece `--compact` y `--limit`, pero ambos son no-op en el caso medido y los hints forman un ciclo. |
+| `AUD-598-B03` / `B-11` | `open` | Persisten 17 anclas de coste con texto `on 5.1.0` dentro de un build 5.8.22; además `risk` no alcanza su ancla publicada en MSAS. |
+| `AUD-598-B04` / `N-02` | `open` parcial | `1e9` y `1_000` son presupuestos válidos, pero se aceptan sin advertir que no vincularán el proceso. |
+
+## Cierres que el artefacto no confirma
+
+La auditoría identifica cinco testigos de `AUD-598-X01`:
+
+- `AUD-597-A01` / `N-01`: cuatro de cinco líneas de `Runs now` siguen sin
+  bounding flag en el wheel.
+- `AUD-596-D01` / `B-13`, `B-14`: `validation` y `data-exposure` carecen de
+  identidad y `schema_version`; además se repite un hash entre repositorios.
+- `AUD-596-D10` / `B-29`: no aparece memoria en cinco payloads completos.
+- `AUD-597-D03` / `N-06`: `route_surface_changed.source_file` es `""` y no
+  lleva `declaration_line`.
+
+La interpretación documentada no es mala fe ni una nueva regresión funcional:
+el ledger y el wheel proceden de estados distintos. La corrección requerida es
+hacer que el build genere y compruebe la relación entre `build_commit` y cada
+commit de cierre antes de publicar.
+
+## Observaciones no convertidas en defectos nuevos
+
+- `ASK-GATE-001`: `migrate-check --ci` y `migrate-apply --ci` siguen rechazados
+  porque son `advisory_only`; la documentación conserva ese precio y no los
+  presenta como gates que puedan fallar.
+- `B-22`: la declaración de `compare` mejoró, pero el residuo de `rc=0` con
+  cero candidatos se mantiene en su fila histórica.
+- `B-21`, `B-26`, `B-27` y `B-30` siguen siendo residuos abiertos ya mapeados a
+  `AUD-596-X02`, `ASK-17`, `C3-128` y la superficie contractual,
+  respectivamente.
+- `N-04` y `ASK-CLI-001` conservan residuos de mensaje/diseño ya trazados en
+  sus filas; no se cierran por la mejora parcial del coste.
+
+## Regla de release
+
+Hasta que `AUD-598-X01` se cierre, una fila no debe marcarse `closed` sólo
+porque el commit de la corrección exista en el checkout. El cierre requiere
+identificar el wheel o artefacto instalado, comprobar que contiene el commit y
+reproducir el comportamiento o declarar explícitamente que se cerró contra un
+fixture distinto del testigo original.

sourcecode-5.8.24/docs/AUDIT-TESTING-CORPUS-2026-08-22.md

@@ -0,0 +1,228 @@
+# Auditoría de features contra el corpus `testing`
+
+**Fecha:** 2026-08-22  
+**Producto auditado:** ASK Engine en el checkout actual  
+**Propósito:** probar las features nuevas de packs, manifests, gates, artefactos,
+MCP y evidencia de despliegue contra repositorios reales, antes de iniciar la
+batería de correcciones.
+
+## Veredicto ejecutivo
+
+La implementación es funcional en el caso normal y no se ha confirmado una
+regresión de análisis, de determinismo ni de los adaptadores MCP/CI. En la
+batería completa, 30 de 37 repositorios produjeron un pack de assessment
+completo (`MEASURED`, seis componentes, `rc=0`). Los siete restantes agotaron el
+presupuesto externo de 120 segundos sin producir el agregado final.
+
+La batería encontró y reprodujo un gap de contrato/ergonomía: usar únicamente
+`--output-dir` dejaba el agregado en stdout. Ese gap queda corregido en el
+checkout actual: el agregado se escribe ahora como `pack.json` (o la extensión
+del formato solicitado), mientras que `--output FILE` explícito conserva
+prioridad.
+
+Este documento no convierte los siete timeouts en una regresión: no existe una
+medición comparable anterior con el mismo corpus, host, flags y presupuesto.
+La repetición completa con snapshots antes/después confirmó que son un límite
+operativo repetible del corpus, no una mutación atribuible a ASK. El baseline
+queda en [`docs/perf/corpus-testing-2026-08-22.md`](perf/corpus-testing-2026-08-22.md).
+
+## Protocolo reproducible
+
+Todas las invocaciones se ejecutaron en modo de sólo lectura:
+
+```text
+ASK_READONLY=1 ASK_TELEMETRY=0 ASK_PROGRESS=off
+ask pack assessment REPO --compact --format json \
+  --output /tmp/assessment.json \
+  --output-dir /tmp/assessment-components
+```
+
+Se usaron procesos aislados, salidas fuera del repositorio y un timeout externo
+de 120 s por repositorio en la batería completa. El timeout es una condición de
+la prueba, no el valor de `ASK_MAX_ANALYSIS_SECONDS` ni una afirmación sobre el
+presupuesto interno de cada comando.
+
+La batería completa cubrió 37 repositorios Git de
+`/Users/user/Documents/workspace/testing`. El corpus no estaba limpio de forma
+homogénea y no se tomó una instantánea de `git status` antes de comenzar; por
+tanto, la prueba no permite atribuir a ASK los cambios ya existentes en esos
+repositorios. Las salidas y resúmenes temporales se conservaron bajo
+`/private/tmp/ask-testing-all*` y `/private/tmp/ask-testing-deep*`, no como
+artefactos del producto.
+
+## Resultado de la batería completa
+
+| Resultado | Repositorios | Interpretación |
+|---|---:|---|
+| Pack completo | 30/37 | `ask-pack-assessment-v1`, `MEASURED`, `rc=0`, seis componentes y agregado parseable |
+| Timeout externo | 7/37 | Sin agregado final dentro de 120 s; requiere baseline y profiling |
+| Traceback observado | 0 | No se observó traceback en los procesos que terminaron ni en la captura de los timeouts |
+| Escritura intencionada | 0 | `ASK_READONLY=1`; la ausencia de snapshot inicial impide certificar causalidad sobre dirty trees |
+
+Los siete repositorios que no terminaron dentro del presupuesto fueron:
+
+- `alfresco-community-repo`
+- `cas`
+- `keycloak`
+- `nacos`
+- `neo4j`
+- `shenyu`
+- `tutorials`
+
+El resultado debe leerse como “30 casos completados y 7 casos no resueltos en
+este presupuesto”, no como “7 fallos funcionales”.
+
+## Muestra profunda corregida
+
+Se repitió una muestra de ocho repositorios con un timeout de 180 s y con
+`--output` explícito para separar el agregado de stdout:
+
+| Repositorio | Tiempo | Resultado |
+|---|---:|---|
+| `spring-petclinic` | 4,85 s | Completo, `MEASURED`, 6 componentes |
+| `examples` | 6,33 s | Completo, `MEASURED`, 6 componentes |
+| `keycloak-config-cli` | 9,07 s | Completo, `MEASURED`, 6 componentes |
+| `mall` | 28,42 s | Completo, `MEASURED`, 6 componentes |
+| `halo` + `--deployment Dockerfile` | 28,83 s | Completo, `MEASURED`, 6 componentes |
+| `open-banking-gateway` + `--deployment docker-compose.yml` | 24,07 s | Completo, `MEASURED`, 6 componentes |
+| `BroadleafCommerce` | 68,47 s | Completo, `MEASURED`, 6 componentes |
+| `spring-security` | 59,12 s | Completo, `MEASURED`, 6 componentes |
+
+En los ocho casos se generaron el agregado y los seis JSON de componente.
+Esto confirma que el problema de la primera pasada de la muestra grande no
+era un fallo del análisis: era la ruta de salida del agregado.
+
+## Hallazgo reproducible: salida agregada con `--output-dir`
+
+### Síntoma
+
+La primera pasada de la muestra profunda usó `--output-dir` sin `--output`.
+Cuatro repositorios terminaron con `rc=1`, error `OUTPUT_TOO_LARGE` y el
+agregado sin persistir:
+
+| Repositorio | Estimación del agregado |
+|---|---:|
+| `mall` | 481.677 tokens / 1.926.709 bytes |
+| `spring-security` | 83.077 tokens / 332.310 bytes |
+| `open-banking-gateway` | 93.781 tokens / 375.127 bytes |
+| `BroadleafCommerce` | 377.907 tokens / 1.511.628 bytes |
+
+El error informa que no hay una variante inline menor y recomienda `--output
+FILE` o `ASK_MAX_OUTPUT_TOKENS=0`. Al repetir con `--output FILE`, los cuatro
+repositorios completaron correctamente.
+
+### Clasificación
+
+`AUD-599-C01` — **P2, gap de contrato/ergonomía, corregido en el checkout actual**.
+
+No era una falsedad del payload ni una pérdida de evidencia: los componentes sí
+se escribían. La implementación ahora hace persistente también el agregado con
+un nombre determinista y mantiene la ruta explícita como autoridad superior.
+La regresión de CLI cubre ambas rutas.
+
+## Contratos comprobados
+
+### Assessment pack
+
+- Los 30 resultados completos publicaron `ask-pack-assessment-v1`.
+- Cada resultado completo publicó seis componentes y fingerprints
+  `content_id`.
+- El catálogo `ask pack list --format json` publicó
+  `ask-pack-catalog-v1`, con los packs `assessment` y `gate`.
+- `ask schema schemas-v1 --format json` resolvió
+  `ask-pack-assessment-v1`, `ask-pack-gate-v1` y
+  `ask-pack-mcp-projection-v1`.
+
+### Gate y códigos de salida
+
+En la muestra profunda, `pack gate REPO --since HEAD~1` produjo
+`ask-pack-gate-v1`, tres artefactos y `verdict=UNVERIFIED`, `exit_code=2`.
+Esto es correcto para una verificación no certificable; no se debe presentar
+como PASS ni como un bug. La muestra no cubrió un caso positivo `PASS`, por lo
+que ese camino queda pendiente de una fixture controlada.
+
+### CI renderers
+
+Los renderers de GitHub Actions y GitLab se probaron en
+`spring-petclinic`, `halo` y `open-banking-gateway`.
+
+- GitHub produjo anotaciones `::warning`/`::notice` válidas y la línea final
+  `ASK_GATE verdict=UNVERIFIED exit_code=2`.
+- GitLab produjo `section_start`, líneas `ASK_GATE_COMPONENT`, línea final
+  `ASK_GATE` y `section_end`.
+
+El chequeo inicial que esperaba literalmente `::group::` era incorrecto: ese
+token no es requisito del renderer y su ausencia no es un defecto.
+
+### MCP
+
+Las llamadas externas sobre `spring-petclinic` devolvieron correctamente:
+
+- `get_assessment_pack`: `success=true`,
+  `ask-pack-assessment-v1`, proyección
+  `ask-pack-mcp-projection-v1`, `MEASURED`.
+- `get_change_safety_gate`: `success=true`, `ask-pack-gate-v1`, la misma
+  proyección, `UNVERIFIED`.
+
+Ambas conservaron `schema_version`, `verdict`, `exit_code` y `manifest`. No se
+observó pérdida de identidad al adaptar el pack a MCP.
+
+### Determinismo
+
+`spring-petclinic` se ejecutó dos veces con los mismos flags. Los seis
+`content_id` fueron idénticos en ambas pasadas:
+
+```text
+migrate-check  ae64790941c201de
+posture        84f0efc7c020d72b
+endpoints      d3ba88844e0f6c9a
+spring-audit   1cf314c5d323438d
+risk           24ce1698282868ce
+audit-report   4109a5829c4472ac
+```
+
+Los manifests completos no fueron byte-identical porque contienen metadatos de
+ejecución (`run_id` y tiempos), lo que es esperado. La identidad del contenido
+sí fue estable.
+
+### Evidencia de despliegue
+
+`Dockerfile` y `docker-compose.yml` se conservaron en `manifest.inputs` en los
+casos de `halo` y `open-banking-gateway`. El payload de posture mantuvo
+honestamente:
+
+```text
+profile_set_is_assumption: true
+decided_in_this_payload: false
+```
+
+Es decir, aceptar el artefacto no implica afirmar qué perfil se ejecuta. No es
+un bug; es la frontera correcta de evidencia cuando el deployment no decide el
+perfil dentro del payload analizado.
+
+## No hallazgos / límites que no deben convertirse en bugs
+
+- No se confirmó regresión en el análisis de endpoints, riesgo, postura,
+  migración o auditoría a partir de esta batería.
+- Los `UNVERIFIED` del gate no son `BLOCK` falsos: no hubo payload suficiente
+  para establecer un bloqueo.
+- La ausencia de un `PASS` en la muestra no demuestra que el camino positivo
+  esté roto; sólo indica cobertura pendiente.
+- Los siete timeouts no son una comparación de regresión hasta obtener baseline
+  anterior con la misma población y protocolo.
+- Los repositorios con dirty trees no permiten atribuir sus cambios a ASK sin
+  un estado inicial capturado.
+
+## Cola para la siguiente fase
+
+| Prioridad | ID | Acción previa a corregir |
+|---|---|---|
+| Cerrado | `AUD-599-C01` | Agregado persistido como `pack.<format>` cuando sólo se usa `--output-dir`; test de fallback, prioridad de `--output` y documentación actualizada |
+| Cerrado como límite operativo | `AUD-599-P01` | Baseline de 37 repositorios y repetición completa; no hay regresión demostrada ni optimización defendible sin timings internos comparables |
+| Cerrado en el checkout actual | `AUD-599-T01` | Fixture controlada para `PASS`, `BLOCK` y `UNVERIFIED`, con códigos de salida diferenciados |
+| Cerrado | `AUD-599-Q01` | Segunda batería completa: 37/37 snapshots Git idénticos antes/después |
+
+Ninguna fila de esta cola debe marcarse como cerrada hasta que exista una
+corrección, una aserción de regresión y una repetición sobre el artefacto que se
+va a distribuir, conforme a la regla de
+[`DEFECT-LEDGER.md`](DEFECT-LEDGER.md).

{sourcecode-5.8.22 → sourcecode-5.8.24}/docs/CACHE.md

@@ -18,12 +18,14 @@ Before optimising a workflow, ask the model that ships with the installed CLI:
 ```bash
 ask cache status .       # what this repository already has
 ask cache model .        # whether a warm helps each command
-ask cache warm .         # compact view plus shared structural layers
-ask cache warm . --agent # add the agent view when that is what the workflow reads
+ask cache warm .                 # every root view plus shared structural layers
+ask cache warm . --views compact # one view only, when that is all the workflow reads
 ```
 
-`cache warm` is not a general acceleration switch. It warms the compact view and shared
-layers for the exact tree and scope you name. A command may receive the finished answer,
+`cache warm` is not a general acceleration switch. It warms every declared view of the
+root analysis and the shared layers for the exact tree and scope you name — each view
+has its own core, so a warm that built one left the others paying a full analysis on
+their first call (AUD-600-R01); the warm names the views it built and any it skipped. A command may receive the finished answer,
 reuse only shared work, or receive no benefit; `cache model` names which. Any analysed-file
 change, committed or not, creates a different cache identity. In CI, persist the ASK cache
 directory only when the job policy allows it; otherwise budget a cold run and use
@@ -79,13 +81,21 @@ the enclosing monorepo, and `cache status` / `cache clear` resolve to that same
 <!-- BEGIN GENERATED: ask cache model --markdown -->
 | Layer | What it stores | What invalidates it | `cache warm` | Written by |
 |---|---|---|---|---|
-| `snapshot` | the rendered answer of the root command (L1 core analysis + L2 view) | tree state · analyzer fingerprint · analysis flags (--env-map, --depth, --docs, --exclude, …) for the core; presentation flags (--compact, --agent, --full, --format, …) for the view | built for the compact view (`--agent` also builds the agent view) | the root analysis alone — `ask <repo>` and `ask cache warm`. AUD-591-Q01: `cache_layers.snapshot` read `cold` in all 240 field invocations across 16 repositories, including after `cache clear --all -y` plus two runs, because every one of those runs was of another command. No other command writes `core-*.json.gz`, so on any run but the root analysis this layer is `cold` by construction and its being cold says nothing about the cache. |
+| `snapshot` | the rendered answer of the root command (L1 core analysis + L2 view) | tree state · analyzer fingerprint · analysis flags (--env-map, --depth, --docs, --exclude, …) for the core; presentation flags (--compact, --agent, --full, --format, …) for the view | built for every declared root view (compact, agent, standard) — one core each, because each view turns a different set of analysers on | the root analysis alone — `ask <repo>` and `ask cache warm`. AUD-591-Q01: `cache_layers.snapshot` read `cold` in all 240 field invocations across 16 repositories, including after `cache clear --all -y` plus two runs, because every one of those runs was of another command. No other command writes `core-*.json.gz`, so on any run but the root analysis this layer is `cold` by construction and its being cold says nothing about the cache. |
 | `ris` | the Repository Intelligence Snapshot — structural index, endpoint index, summaries | tree state (the snapshot records the tree it describes) | rebuilt on every warm | every command that reads it |
 | `cir` | the shared Canonical IR — the Java parse every knowledge command reuses | tree state · analyzer fingerprint · schema version | built | every command that reads it |
 | `task` | one `prepare-context` task answer, per task and per option set | tree state · task · --symptom / --all / --include-config / --format | never — a warm does not run any task | every command that reads it |
 | `timeline-samples` | one watched metric as measured at one sampled commit, so a series re-run over an overlapping range measures only what is new (C3-129) | the analyzer fingerprint and the commit sha, both in the key. The tree at a sha never changes, so an entry is never stale — the fingerprint is what keeps a corrected analysis from being served from a previous release. `timeline --no-cache` (or `ASK_TIMELINE_NO_SAMPLE_CACHE=1`) measures every sample again | never — a warm analyses the working tree, not history | every command that reads it |
 | `parse` | the symbol extraction of one file, keyed by its bytes | the file's own bytes and the extractor's source — content-addressed, so it is never stale and never needs invalidating. Bounded by `ASK_PARSE_CACHE_MAX_MB` (default 512 MB, least-recently-used, swept every 32 MB written rather than on every write, and the last interval before a process exits is never swept — so the store rests up to one interval per writing process above its budget, and the figure to size a disk by is the ceiling `cache status` publishes beside it, not the budget itself (ASK-15)) and reported in `cache status` (F-AR) | filled for every Java file the warm parses | every command that reads it |
 
+| Root view | Invocation | Its own core | `cache warm` |
+|---|---|---|---|
+| `compact` | `ask <repo> --compact` | turns dependencies, env map, code notes and architecture on; bounded, and the view the front page recommends first | builds it |
+| `agent` | `ask <repo> --agent` | everything the compact core has plus the module import graph, and it drops the file tree — a different core, not a different rendering. This is the view the root help recommends to AI agents | builds it |
+| `standard` | `ask <repo>  (and `ask <repo> --full`, which is this view with the display caps raised)` | the raw contract view: none of the compact analysers run, so its core is neither of the two above and a warm of either leaves it cold | builds it |
+
+Each view turns a different set of analysers on, so each has its own L1 core: a warm that builds one leaves the others paying a full analysis on their first call (AUD-600-R01).
+
 Measured on BroadleafCommerce (2 000+ Java files, dirty tree), warm machine, 3.2.2, each command in isolation.
 
 Measured on release 3.2.2. Figures are that build's on that repository, not a prediction for yours; the nightly performance gate (`docs/perf/REGRESSION-GATE.md`) is what re-measures them.
@@ -96,7 +106,7 @@ The second anchor is the field one, and it carries its own build for the same re
 |---|---|---|---|---|---|
 | `ask (root)` | the answer | --compact 13.3 s cold → 0.3 s warm (cold re-measured on 3.7.0: was 19.3 s, C3-6); --agent --full --env-map --depth 20 34.7 s → 33.9 s (no gain) | yes | `snapshot`, `ris`, `parse` | `--compact` is what a warm stores by default; `--agent` needs `cache warm --agent`. `--env-map`, `--depth N` and `--exclude` change the *analysis*, so they miss the warmed core and rescan — the 171 s the field measured after a 103 s warm on 4.10.4. Eval #23 timed the same invocation at 72,3 s from a purged cache, and then paid 71,7 s again for `--agent` on the state it had just analysed (C3-103). |
 | `posture` | the shared work | 10.1 s → 1.6 s | no | `cir`, `parse` | Resolves the conditional bean graph on every run, over the shared CIR a warm builds — the parse it used to repeat for itself. `--diff` compares two profile sets over that one IR, so the second side costs the resolution only. |
-| `risk` | the shared work | not measured on the battery yet — the composition is bounded by the `spring-audit` + `impact-chain` costs listed here, not by new analysis | no | `cir`, `parse` | Composes what the audit, impact-chain and the posture already answer, so it pays each of their costs once over the shared CIR a warm builds — one parse for the whole composition, and the reachability query is cached per symbol within the run. |
+| `risk` | the shared work | 240.1 s on the audited monolith (13th pass, rc 0, 178 171 B). The two rounds before it exhausted every budget they gave it — 301 s under a 300 s budget (5.8.20) and 401 s under 400 s (5.8.22) — and the refusal with its budget is the honest anchor for a run that does not finish, never a smaller number measured somewhere else | no | `cir`, `parse` | Composes what the audit, impact-chain and the posture already answer, so it pays each of their costs once over the shared CIR a warm builds — one parse for the whole composition, and the reachability query is cached per symbol within the run. |
 | `enrich` | the shared work | not measured on the battery yet — bounded by the `risk` composition, plus reading one JSON file | no | `cir`, `parse` | Runs the same composition as `risk` over the repository, then joins a SARIF log to it. Reading the log is negligible; everything a warm helps with is the repository side, so what it buys is what it buys `risk`. |
 | `audit-report` | the shared work | not measured on the battery yet — bounded by `risk` + `posture`, or by `posture` when `--from-risk` is supplied | no | `cir`, `parse` | Packages `risk` and `posture` evidence. A warm helps the repository side; `--from-risk risk.json` skips risk recomputation entirely and only builds the posture/report projection. |
 | `migrate-recipe` | the shared work | not measured on the battery yet — bounded by `migrate-check` on the same repository (openmrs-core ~2 s) | no | `parse` | Runs the same scan as `migrate-check` and projects its findings into an OpenRewrite recipe, so it buys exactly what a warm buys `migrate-check`: the parse, not the rule pass. |

{sourcecode-5.8.22 → sourcecode-5.8.24}/docs/COMMERCIAL-OFFER.md

@@ -17,6 +17,10 @@ and `ask pack gate` for recurring change safety in CI. See
 [`PRODUCT-ROUTE.md`](PRODUCT-ROUTE.md) for the normative scope and evidence
 contract.
 
+The assessment can be delivered as JSON evidence for automation or as an
+executive/procurement Markdown report with `--format markdown`; both views are
+renderings of the same measured pack.
+
 The gating axis is **size + automation**, never capability. Nothing is crippled on Free;
 you pay exactly when the tool is replacing the most engineering time — a monolith above
 **500 Java source files**, or CI/CD automation. **€19/mo · €190/yr, per developer.**