@andresmassello/uscha 1.51.1 → 1.51.3

package/uscha-kit/CHANGELOG-1.37.0.md

@@ -1,41 +0,0 @@
-# CHANGELOG 1.37.0 ? ADR experiments: visible hypotheses
-
-Decision #10 from the deferred analysis is implemented narrowly: an ADR can now say
-`Status: Experiment` when the decision is a bounded hypothesis that needs real feedback.
-
-## What changed
-
-- `dashboard --json` now preserves both ADR layers:
-  - legacy `status` remains the coarse Mirador bucket (`done` / `prog` / `todo`);
-  - `adr_status` preserves the authored ADR status (`accepted`, `proposed`, `experiment`, etc.).
-- Experimental ADRs carry advisory metadata:
-  - `experiment_valid`
-  - `experiment_missing`
-  - `review_by`
-  - `review_trigger`
-  - `feedback_signal`
-  - `expired`
-- Top-level `adr_experiments` summarizes `open`, `malformed`, `expired`, and `ids`.
-- Mirador renders experiment ADRs as `experiment`, `experiment invalid`, or `experiment overdue`.
-- `Status: Experiment` is intentionally **not** readiness scoring and **not** a hard PR gate.
-
-## Required sections for `Status: Experiment`
-
-An experiment ADR must include:
-
-- `Hypothesis`
-- `Feedback Signal`
-- `Review By` or `Review Trigger`
-- `Promote Criteria`
-- `Rollback / Supersede Criteria`
-
-Missing/expired metadata is visible/advisory so the operator does not confuse a hypothesis with
-an accepted decision.
-
-## Smoke
-
-- Full smoke: `209 ok ? 0 fail`.
-- New T62 covers:
-  - valid experiment ADR exposed in `dashboard --json`;
-  - malformed/expired experiment counted as advisory;
-  - Mirador renders `experiment` without changing readiness.

package/uscha-kit/CHANGELOG-1.38.0.md

@@ -1,11 +0,0 @@
-# uscha-kit 1.38.0 ? contract closure loop
-
-## Added
-
-- `spec-change-request`: records a human contract-change bridge from evidence/doubt to SPEC/ADR amendment, surfaces it in `readiness --json`, and derives `phase=escalated` while open.
-- `golden-diff --labels golden-labels.json --json`: reports approved fixture intent as `intended`, `observed-accidental`, or `unknown` without weakening byte comparison.
-- `summary --json` now includes `post_merge_calibration` counts for production findings, SPEC doubts, and SPEC change requests so retros can calibrate the method from real reopen signals.
-
-## Verification
-
-- Smoke target: 216/216 green.

package/uscha-kit/CHANGELOG-1.39.0.md

@@ -1,14 +0,0 @@
-# uscha-kit 1.39.0 ? universal machine installer
-
-## Added
-
-- `install-uscha.py`: one stdlib machine installer for Codex and Claude.
-  - Codex target creates a personal local plugin at `~/plugins/uscha` and updates `~/.agents/plugins/marketplace.json`.
-  - Claude target installs global skills/hooks under `~/.claude`.
-  - Supports `install`, `doctor`, `version`, `init`, `--dry-run`, testable `--home`, and `--mode copy|link`.
-- Codex plugin manifest at `.codex-plugin/plugin.json` for plugin-first adoption.
-- Skill fallback examples now include Codex plugin/raw-skill install paths, not only `~/.claude`.
-
-## Verification
-
-- Smoke target: 222/222 green.

package/uscha-kit/CHANGELOG-1.4.0.md

@@ -1,68 +0,0 @@
-# dev-loop-kit 1.4.0 — adapter `type: python` (2026-07-02)
-
-El desbloqueante del dogfooding: el engine ahora mide repos Python (pytest + ruff +
-mypy + coverage.py). Diseño: `docs/HANDOFF_python-adapter_dev-loop-kit.md` +
-`docs/HANDOFF_python-adapter_ADDENDUM.md` (los 2 bugs A1/A2 de la revisión están
-incorporados). Smoke suite: 24/24 (17 existentes maven/flutter intactos + 7 python).
-
-## Engine (qa_ledger.py)
-
-- **`python_coverage()`** — Cobertura `coverage.xml` (`pytest --cov --cov-report=xml`),
-  atributos `lines-covered`/`lines-valid`; primer match gana (`coverage.xml`,
-  `reports/coverage.xml`). Reporte ausente o ilegible ⇒ `report_found: False`
-  (UNMEASURED — jamás un número inventado).
-- **`python_test_count()`** — JUnit XML de `pytest --junitxml`. **Maneja el root
-  ENVUELTO** (`<testsuites><testsuite …>`, familia xunit2 del pytest moderno) además
-  del root plano — leer atributos del wrapper contaría 0 en silencio (ADDENDUM A2).
-- **`parse_ruff()`** — `ruff check --output-format=json`; severidades: `S<dígito>`
-  (bandit/security, anclado para no pisar SIM/SLF/SLOT) y `E9*`/`F82*` (errores
-  reales) → HIGH, `B<dígito>` (bugbear) → MEDIUM, resto → LOW. **`code: null`**
-  (syntax error del ruff moderno ≥0.5) → HIGH siempre — sin esto, la rotura real
-  pasaba el gate como LOW en silencio (hallazgo del review de contexto fresco).
-- **`parse_mypy()`** — salida de texto (`file:line: error: msg [code]`);
-  error → HIGH, warning → MEDIUM, note → INFO.
-- **`ingest-gate`** — branch por `type` del repo: python busca `reports/ruff.json` /
-  `reports/mypy.txt` (o `--ruff`/`--mypy` explícitos); combinado se loguea como
-  `python-qa-gate`. Mismo contrato de ausencia que Java.
-- **Guard UNMEASURED ampliado** (ADDENDUM A1): `static_unmeasured` ahora aplica a
-  `maven` Y `python` — un repo python cuyos linters nunca corrieron puntúa 0.0 en la
-  dimensión static, no 1.0 (habría reintroducido "el silencio es éxito" justo en el
-  repo del dogfooding).
-- **LOC python**: test = `tests/`/`test/` en el path o `test_*.py`/`*_test.py`;
-  prod = el resto de los `.py` (src layout o paquete raíz). `SKIP_DIRS` suma
-  `.venv`, `venv`, `.tox`, `__pycache__`, `.mypy_cache`, `.ruff_cache`,
-  `.pytest_cache` (sin esto, el LOC contaría el virtualenv entero).
-- `_rel_src()` reconoce paths repo-relativos `src/...` (como los emiten ruff/mypy)
-  para IDs de finding estables. **Nota de migración**: si un linter Java emitía paths
-  relativos `src/...` (poco común — checkstyle/PMD suelen emitir absolutos), sus
-  finding-IDs cambian en 1.4.0; un ledger no debería cruzar el upgrade — re-baselinear
-  con un ingest fresco después de actualizar.
-- `SOURCE_EXT`: `"python": {".py"}`; `.py` también en `generic`.
-
-## Ya cubierto sin cambios (ADDENDUM A3)
-
-- simplicity-check: `.py` ya estaba en `_SIMPLICITY_CODE_EXT`.
-- gate-check: ya reconocía `test_*.py`, `def test_`, `# noqa`, `# type: ignore`.
-
-## Config / docs
-
-- `dev-loop.config.json`: `test_command_python` + repo de ejemplo `data-lib`; version 1.4.0.
-- Kit README: tabla de reportes Python; SKILL.md y docs actualizan
-  `maven|flutter` → `maven|flutter|python`; refs de versión vigente → 1.4.0 (ES+EN).
-
-## Smoke (tests/smoke-engine.sh)
-
-- `repo-c` (`type: python`) con fixtures sintéticos: Cobertura 8/10 → 80.0%,
-  junit ENVUELTO → 5 tests, LOC prod=3/test=2, ruff 2 findings (1 gateado),
-  mypy 1 error (gateado), UNMEASURED pre-ingest → medido post-ingest.
-
-## Criterio pendiente (fase piloto)
-
-Criterio 2 del HANDOFF — dry-run de solo lectura contra el repo piloto real
-(coverage ±0.1 vs pytest-cov, test count exacto, ruff/mypy en 0 sin UNMEASURED).
-Se corre desde la sesión del proyecto piloto consumiendo este kit.
-
-## Diferido consciente (sin cambios desde 1.3.0)
-
-- rebuild: densidad de asserts por test-file en la firma.
-- Perfiles A-E mecanizados (`--profile`).

package/uscha-kit/CHANGELOG-1.40.0.md

@@ -1,16 +0,0 @@
-# uscha-kit 1.40.0
-
-## Added
-
-- Added npm/npx package router (`@andresmassello/uscha`) with `uscha` and `uscha-kit` CLI bins.
-- Added smoke coverage for the Node router delegating to the canonical Python installer.
-
-## Changed
-
-- Documented npm-first install commands for public adoption while keeping `install-uscha.py` as the source of truth.
-
-## Verification
-
-- `node bin/uscha.js version --json`
-- `npm pack --dry-run --json`
-- `bash uscha-kit/tests/smoke-engine.sh`

package/uscha-kit/CHANGELOG-1.40.1.md

@@ -1,11 +0,0 @@
-# uscha-kit 1.40.1
-
-## Fixed
-
-- Exclude Python bytecode artifacts (`__pycache__`, `*.pyc`, `*.pyo`) from the npm package tarball.
-- Strengthen the npm smoke check so package dry-run fails if Python bytecode artifacts would be published.
-
-## Verification
-
-- `npm publish --dry-run --access public`
-- `bash uscha-kit/tests/smoke-engine.sh`

package/uscha-kit/CHANGELOG-1.40.2.md

@@ -1,13 +0,0 @@
-# uscha-kit 1.40.2
-
-## Changed
-
-- Reworked install documentation around the npm/npx path for Codex, Claude Code, and mixed-machine setups.
-- Added `uscha-kit/INSTALL.md` as the focused install guide.
-- Kept Git checkout/link mode, Claude Code plugin commands, and manual copy as secondary/debugging options with tradeoffs.
-
-## Verification
-
-- `npm view @andresmassello/uscha version`
-- `npx --yes @andresmassello/uscha@latest version --json`
-- `bash uscha-kit/tests/smoke-engine.sh`

package/uscha-kit/CHANGELOG-1.41.0.md

@@ -1,18 +0,0 @@
-# uscha-kit 1.41.0 — safety release
-
-Significant backwards-compatible release covering WU1–WU7.
-
-- **WU1:** measured `pr-ready`, fail-closed Ruff/JUnit schemas, and validated counters.
-- **WU2:** safe init, authoritative doctor, portable active Claude hook, marketplace preflight, and staged rollback.
-- **WU3:** evidence-bearing PF/SD/SCR closure, config validation, and monotonic iterations/counters.
-- **WU4:** integration readiness heterogeneous-event false-green prevention, unknown escalation repo rejection, and Spanish ADR label folding.
-- **WU5:** ADR templates use AC IDs; the resolved constitution blocker example includes `--escape-analysis`.
-- **WU6:** npm router probes usable Python >=3.8 before one installer invocation with no retry; workbench-doctor uses portable Python and the current skill roster.
-- **WU7:** complete Claude installation transaction with rollback across managed skills, hook, settings, and marker; T77 covers existing and absent marker rollback.
-- **P0 hardening:** Mirador script/DOM injection removed; structured static reports including Clippy fail closed; stale JUnit evidence cannot satisfy `pr-ready`.
-- **P1 installer hardening:** init preflights all targets and rejects symlinks; Codex restores marketplace state during rollback; hook matcher and malformed Claude settings are validated.
-- **P1 ledger hardening:** readiness configuration is validated completely and PF/SD/SCR closure is immutable after its first resolution.
-- **Documentation truth-pass:** exact 29-command reference, npx-first Codex/Claude adoption, eight-skill inventory, completed EN twins, and refreshed diagrams.
-- Smoke suite: **351/351 green**.
-
-Known boundary: Codex and Claude targets are independently transactional; `--target both` remains sequential rather than cross-target atomic.

package/uscha-kit/CHANGELOG-1.41.1.md

@@ -1,53 +0,0 @@
-# uscha-kit 1.41.1 — adversarial-review hardening (2026-07-11)
-
-An independent adversarial review of the 1.41.0 "safety hardening" release (4 parallel
-reviewers, each tasked to REFUTE the safety claims) confirmed three real defects — two of
-them in the engine's core "measured beats narrated" promise. Fixed here, each with a
-regression check that fails without the fix. Smoke suite: 355/355.
-
-## Fixes
-
-### 1 — `pr-ready` accepted a JUnit that LIES (HIGH, core-promise)
-`_junit_counts` read `tests`/`failures`/`errors` from the `<testsuite>` **attributes** and
-never reconciled them against the actual `<testcase>`/`<failure>`/`<error>` **elements**. A
-report declaring `failures="0"` while containing a real `<failure>` element read as all-green
-and could satisfy `pr-ready`. Fix: honor the child elements, fail-closed —
-`failures = max(attr, element_failures)`, same for errors. A present failure can no longer be
-attribute-declared away. Attribute-only summary suites (no `<testcase>` elements — the form
-many emitters legitimately use) keep their counts, so real adapters are unaffected.
-`qa_ledger.py:_junit_counts`. Regression: smoke **T69b**.
-
-### 2 — integration readiness trusted the single last event (MEDIUM-HIGH, core-promise)
-The integration dimension read `integ_steps[-1]`, so a trailing green test-only step masked
-an earlier FAILING integration gate (dim → 1.0). Fix: green now requires **0 open gated
-findings across the latest record per integration tool** AND the latest test event passing —
-so a failing gate is not masked, while a same-tool re-run that clears the gate is still seen.
-`qa_ledger.py:cmd_readiness`. Regression: smoke **T69c**. (The 1.41.0 CHANGELOG claimed WU4
-fixed this; the code still used `integ_steps[-1]` — the claim was inaccurate.)
-
-### 3 — Codex install rollback destroyed the pre-existing plugin (HIGH, data loss)
-`install_codex` gated its restore on `swapped` (only true after BOTH the backup-move and the
-install-move succeeded), then the `finally` deleted the backup unconditionally. If the
-`stage → plugin_root` swap failed after the original was moved to backup (a real Windows AV /
-locked-handle failure on reinstall), the restore was skipped and the `finally` deleted the
-only surviving copy — silently. Fix: gate the restore on the **backup existing** (not on
-`swapped`), and drop the backup only on SUCCESS (never in `finally`), so a failure restores
-the original and a hard interrupt leaves it intact. Mirrors the Claude path, which already did
-this. `install-uscha.py:install_codex`. Regression: smoke **T77b**. (This is exactly the
-"rollback leaves partial state" category 1.41.0 claimed to close — it closed it for Claude,
-not Codex.)
-
-## Honest limits (not fixed here, by design)
-- A JUnit `<testsuite tests="3"/>` with ZERO `<testcase>` elements is still credited as 3
-  tests. This is the same trust the engine extends to any tool-emitted summary (a coverage %
-  in a coverage report), and closing it would reject the legitimate attribute-only reports
-  many emitters produce. The defense against a **forged** bare report is the "evidence
-  captured by execution" invariant (the SKILL runs the real test command; the agent does not
-  hand-write the XML) + the human merge gate — not per-report forgery detection.
-
-## Also
-- `docs/skills-referencia.html` + `-EN.html`: stale "24 subcommands" → 29 (truth-pass).
-
-## Note on the test suite
-The installer/npm smoke checks hardcode the version string, so a version bump requires
-updating those literals (done here). A follow-up could have them read `VERSION` dynamically.

package/uscha-kit/CHANGELOG-1.41.2.md

@@ -1,34 +0,0 @@
-# uscha-kit 1.41.2 — mirador launch UX (2026-07-18)
-
-A usability release for the mirador (the bird's-eye dashboard). No engine measurement
-logic changed; this makes the mirador honest about "not ready" and trivial to launch from
-inside a running project — "nada de magia oculta". Smoke suite: 357/357.
-
-## Fixes
-
-### 1 — "NOT READY" no longer reads as "hasn't started" (readiness title)
-The mirador mapped the `NOT READY` band (score 0–49) to a fixed title `"Todavia no arranca"`
-("hasn't started yet"). But a project at, say, readiness 22 HAS started — it simply lacks
-enough **measured** evidence. The wording contradicted the score right next to it. Fix: the
-`NOT READY` title is now score-aware — `"Todavia sin evidencia medida"` only at score 0, and
-`"En construccion -- evidencia insuficiente"` above 0. The dead fixed entry was removed from
-`_MIRADOR_TITLE` so the old wording cannot reappear. `qa_ledger.py:cmd_dashboard`.
-Regression: smoke **T78**.
-
-## Launch UX
-
-### 2 — `mirador-render.py` self-resolves its siblings
-`--engine` (the `qa_ledger.py` path) and `--template` (the mirador HTML) now default to the
-renderer's **sibling** skill files. From any project you can run the renderer — or
-`/uscha-mirador` — with just `--ledger`, no long absolute paths. `mirador-render.py:main`.
-
-### 3 — the rendered mirador opens itself, and always prints where it is
-After writing `mirador.html` the renderer opens it in the default browser (best-effort:
-`os.startfile` on Windows, `open`/`xdg-open` elsewhere — never fails on headless/CI) and
-prints the **absolute path** on an `OPEN IT:` line. A new `--no-open` flag suppresses the
-auto-open; the live-watch loop (`mirador-watch.sh`/`.ps1`) and the smoke suite pass it so
-they never spawn browser tabs on every refresh. `mirador-render.py`.
-
-## Note on the test suite
-The installer/npm smoke checks still hardcode the version string, so this bump updated those
-literals (T44, T66/T67 assertions). A follow-up could have them read `VERSION` dynamically.

package/uscha-kit/CHANGELOG-1.41.3.md

@@ -1,30 +0,0 @@
-# uscha-kit 1.41.3 — mirador live-view no longer spams browser tabs (2026-07-18)
-
-A hotfix for a regression shipped in 1.41.2. Smoke suite: 358/358.
-
-## Fix
-
-### mirador-render auto-open + `--refresh` spawned a new browser tab every cycle (HIGH, regression)
-1.41.2 made `mirador-render.py` auto-open the rendered file in the browser. But the live
-second-screen view (`mirador-watch`, and any render with `--refresh N`) injects a
-`<meta http-equiv="refresh">` so a single open tab **reloads itself** every N seconds — and
-the renderer is re-invoked each cycle. Auto-opening on every invocation therefore spawned a
-**new browser tab on every refresh** (30s → a wall of tabs), making the machine unusable while
-a watch loop ran. The `--no-open` flag was not enough on its own: a watch process started
-before 1.41.2 calls the renderer without it.
-
-Fix: auto-open is now suppressed whenever `--refresh > 0` (live mode) — the meta-refresh owns
-the reload, in one tab. Auto-open remains for a genuine **one-shot** render (`/uscha-mirador`
-with no `--refresh`), and `--no-open` still suppresses it everywhere. The `OPEN IT: <abs path>`
-line is always printed, so the live view is one manual open away and never hidden.
-`mirador-render.py:main`. Regression: smoke **T79** (live → 0 opens, one-shot → 1, `--no-open`
-→ 0).
-
-Because the renderer is a fresh subprocess each cycle, a running watch loop picks up this fix
-on its **next** render with no restart — and for a link/junction skill install (the renderer
-lives in this repo), no reinstall either. A copy-mode install needs a reinstall to get the
-fixed file.
-
-## Note on the test suite
-The installer/npm smoke checks still hardcode the version string, so this bump updated those
-literals. A follow-up could have them read `VERSION` dynamically.

package/uscha-kit/CHANGELOG-1.42.0.md

@@ -1,41 +0,0 @@
-# uscha-kit 1.42.0 — the mirador answers "how's it going / what's blocking / what's next" (2026-07-18)
-
-Until now the mirador showed the project's SKELETON — the phase trail, the counters, the
-invariant names — but a real early-stage project rendered as a wall of empty-looking cards
-that left you with "sabor a nada": you could not tell, at a glance, how it was going, what was
-blocking it, or what to do next. This release makes the mirador lead with exactly those three
-answers — and every one of them is **measured**, not narrated. Smoke suite: 359/359.
-
-## The status story (hero block)
-
-A three-line block sits under the verdict, each line derived purely from fields the engine
-already emits in `dashboard --json` (no new engine logic, still model-agnostic):
-
-- **Cómo viene** — the measured sub-scores, verbatim: e.g. `coverage 0% · simplicity FAIL`.
-- **Qué lo traba** — the hard blockers, each traceable to a measured field: failing gates
-  (sub-scores reading FAIL/MISS), the readiness cap (`readiness.sub`), open discovery intake
-  (production findings / spec doubts / spec change requests), and expired/malformed ADR
-  experiments. A failing gate is NOT double-reported as its flipped CONSTITUTION invariant.
-  When nothing blocks, it says so and names the dimensions still lacking measured evidence.
-- **Qué sigue** — the active phase in the derived FSM plus its method and execution hint
-  (`tier` / `model` / `effort`), so the operator sees the next concrete move.
-
-`mirador.template.html:renderStatus`. Regression: smoke **T80**.
-
-## Empty cards earn their place
-
-Cards no longer render as dead space when a real project has nothing for them yet:
-
-- **The 6 layers of truth** (`capas`) — the engine never feeds this for a real project (it was
-  demo-only), so the whole card now HIDES when empty instead of showing an empty grid.
-- **Specifications / ADRs / QA loops** — when empty, show a one-line hint of what fills them
-  and in which phase, instead of a blank list or a bare `0/0`.
-
-## Safety
-All status/hint DOM is built with `textContent` / `createTextNode` (no `innerHTML`), so the
-1.41.0 Mirador XSS hardening still holds — smoke P0-A (script-context escaping, no HTML sinks)
-stays green.
-
-## Note on the test suite
-The installer/npm smoke checks still hardcode the version string, so this bump updated those
-literals. A follow-up could have them read `VERSION` dynamically.

package/uscha-kit/CHANGELOG-1.43.0.md

@@ -1,37 +0,0 @@
-# uscha-kit 1.43.0 — `uscha mirador`: one command, no python, no paths (2026-07-18)
-
-Adoption fix. Bringing up the mirador for a running project meant typing
-`python <long-skill-path>/mirador-render.py --ledger QA-LEDGER.json` — friction that pushed
-users away. The kit already had an npm/npx router (`bin/uscha.js`) forwarding to
-`install-uscha.py`; it was missing one verb. Smoke suite: 361/361.
-
-## New: the `mirador` verb
-
-From the root of any project that has a `QA-LEDGER.json`:
-
-```
-uscha mirador                        # render + open the dashboard
-uscha mirador --watch                # live second-screen view (auto-refresh, one tab)
-npx @andresmassello/uscha mirador    # same, zero install
-```
-
-No python, no paths, no flags required. The verb:
-- resolves the renderer inside the kit (either skill-tree layout), which self-resolves its own
-  engine + template siblings;
-- defaults the ledger to the `QA-LEDGER.json` convention in the current directory
-  (`--ledger` to point elsewhere);
-- one-shot renders and opens the file; `--watch` opens ONE self-reloading tab and re-renders
-  every `--interval` seconds (default 30) WITHOUT re-opening — no browser-tab spam (honors the
-  1.41.3 live-mode rule);
-- fails clearly (exit 1) when the ledger is missing, instead of a stack trace.
-
-`install-uscha.py:cmd_mirador` (exposed through the existing `bin/uscha.js` router).
-Regression: smoke **T81**.
-
-## Docs
-`INSTALL.md` gains a "See the dashboard (mirador)" section; the uscha-mirador `SKILL.md`
-leads its human/terminal path with `uscha mirador`.
-
-## Note on the test suite
-The installer/npm smoke checks still hardcode the version string, so this bump updated those
-literals. A follow-up could have them read `VERSION` dynamically.

package/uscha-kit/CHANGELOG-1.44.0.md

@@ -1,90 +0,0 @@
-# uscha-kit 1.44.0 — brownfield truth: what the engine could not measure, it now says (2026-07-21)
-
-Field release. Everything here comes from a retrospective of a real run: a gzip migration
-across two legacy Java 8 / Ant repositories with fiscal data, taken end to end
-(reverse-discovery → approved golden → adr-refine → devloop → merge). The run succeeded, and
-it surfaced four defects that all share one root: **the engine knew something and did not say
-it**. Smoke suite: 366/366.
-
-Not in this release: the inert `risk_profile` (declared in config, read by nobody) — the
-deepest finding of the same retro. It gets its own release; see the note at the end.
-
-## Fixes
-
-### 1 — coverage: "no report at all" is not the same fact as "a measured 0%"
-`static_unmeasured` already taught the engine that *silence is not success* for the static
-gate. Coverage had no equivalent: a repo with no instrumentation scored exactly like a repo
-whose report says 0%, and the operator had no way to tell which — the field report had to read
-`qa_ledger.py` to find out. Two legacy repos with 18 green tests sat at readiness **66.7**
-purely because of this, an artificial ceiling in exactly the brownfield scenario
-`reverse-discovery` targets.
-
-Fix: the coverage parsers already returned `report_found`; the readiness dimension now uses it.
-`coverage_unmeasured` is exposed per repo (`facts`) and aggregated
-(`facts.coverage_unmeasured_repos`), and readiness prints the repos with the remedy named.
-
-The score is deliberately **unchanged** (still 0.0): auto-redistributing the weight would let
-any repo raise its score by DELETING its instrumentation — the precise anti-pattern
-`static_unmeasured` exists to prevent. The relief is a **human declaration**:
-`defaults.readiness_weights.coverage = 0` redistributes the weight, is versioned in the config,
-and carries provenance (config = requirement, kit 1.17.0). Silence never buys points; a
-declaration does. `qa_ledger.py`. Regression: smoke **T82** (pins the field's 66.7 → 100.0).
-
-### 2 — no `ant` project type
-Supported types were maven, gradle, flutter, python, node, go, rust, dotnet, cpp, swift. Ant —
-the dominant build tool in legacy Java, i.e. the terrain the brownfield front is FOR — was
-missing, so the field run had to declare two Ant projects as `maven`. Added `ant`: because Ant
-has no standard output layout (the `todir` is build-file defined), its reports are discovered
-**recursively by name** (`**/TEST-*.xml`, `**/jacoco.xml`) instead of guessing one convention.
-Also added to the lint-capable set for `static_unmeasured`.
-
-Recursive discovery needs two guards, both found by an adversarial review of this very
-release: (a) reports are pruned of **third-party trees only** (`node_modules`, `vendor`,
-`.git`, `.venv`...) — deliberately NOT of `build/`, `target/` or `coverage/`, because reports
-LIVE in build output and the existing `SKIP_DIRS` (built for scanning SOURCE) would have
-hidden exactly the files being looked for; and (b) name-based discovery is **tolerant**: a
-`TEST-*.xml` that turns out not to be a JUnit `<testsuite>` — or that has a valid root but
-impossible counters, e.g. a truncated run — is dropped instead of aborting the whole run with
-`SystemExit(2)`. Precise patterns (maven/gradle) keep the hard failure.
-
-Crucially, a drop is **never silent**: `report_found` now means a *usable* report, and every
-dropped file is returned in `skipped_reports`, persisted in the snapshot, exposed per repo
-(`facts.dropped_reports`) and aggregated (`facts.dropped_report_repos`), and named by
-`readiness` on **stdout** — because `dashboard --json` (the mirador) captures stdout, so a
-stderr-only warning would be invisible exactly where operators look. Dropping evidence quietly
-would be silence buying a pass, which is the one thing this engine must not do.
-Regression: smoke **T83**.
-
-**Known limitation (by design, for now):** because report discovery for `ant` searches by name
-and deliberately does not prune `build/`/`target/` (reports live there), a nested sibling Maven
-or Gradle module inside an `ant` repo will have its reports swept into the parent's totals.
-That is the trade for a build tool with no convention; declare such modules as their own repos
-in `uscha.config.json` to keep their evidence separate.
-
-### 3 — JUnit discovery too narrow (an integrity issue, not a convenience one)
-The engine looked for `target/surefire-reports/TEST-*.xml`, `reports/junit.xml` and
-`junit.xml`. A runner writing a DIRECTORY of per-class XML under `reports/junit/` was invisible,
-so the operator **hand-copied reports** to a path the engine knew. That is not friction: a
-hand-placed report breaks the *evidence captured by execution* invariant, which is the whole
-defense against a forged report. Now `reports/junit/**/*.xml` is discovered too.
-Regression: smoke **T83**.
-
-### 4 — a Windows reserved filename took down a whole gate
-`waste-check` died with `ValueError: path is on mount '\\.\nul'` on a repo containing a file
-named `nul`; the gate simply never ran. Reserved device names (`nul`, `con`, `aux`, `prn`,
-`com1..9`, `lpt1..9`) are now skipped in all four tree walks. Detection splits on the FIRST
-dot, not the last: Windows treats `nul.tar.gz` as the device too, and an initial fix using
-`os.path.splitext` left that crash reachable (`_newest_source`, `count_loc`,
-`_test_file_set`, `_waste_repo_hashes`). A weird filename degrades to a skip, never a crash.
-Regression: smoke **T84** (multi-extension cases, and no false positives: `nullable.py`,
-`console.ts`, `com1x.go`).
-
-## Note on the risk profile
-The same retro verified that `risk_profile` appears **zero times** in the engine: it is declared
-and read by nobody, so the decision of "how much process does this change deserve" rests
-entirely on the operator. That is the difference between a methodology and a convention, and it
-is too deep to bundle here — it is the next release.
-
-## Note on the test suite
-The installer/npm smoke checks still hardcode the version string, so this bump updated those
-literals.

package/uscha-kit/CHANGELOG-1.44.1.md

@@ -1,26 +0,0 @@
-# uscha-kit 1.44.1 — `uscha init` is per-file, not all-or-nothing (2026-07-21)
-
-Found by dogfooding: running `uscha init` on the kit's own repository. Smoke suite: 367/367.
-
-## Fix
-
-### `init` was unusable on any repo that already had a `CLAUDE.md`
-`cmd_init` copies four files (`uscha.config.json`, `CLAUDE.md`, `CONSTITUTION.md`,
-`.gitattributes`). It was **all-or-nothing**: if any target already existed and differed from
-the template, the whole init aborted and wrote **nothing**. But `CLAUDE.md` differs in every
-repo that already uses Claude Code — i.e. exactly the repos adopting uscha — so init reliably
-did nothing, and the only escape was `--force`, which would **overwrite the user's own
-`CLAUDE.md`**. The first command of the self-application run hit this immediately.
-
-Fix: `init` is now per-file. The non-conflicting files are written regardless; each conflict is
-reported and left **untouched** (resolve by hand, or re-run with `--force` to replace it
-deliberately). Status is `partial` when some files were written and conflicts remain, and the
-exit code stays **1** while any conflict is pending, so the partial state is visible and
-scriptable. A clean repo still initializes fully and exits 0 — backward compatible. The JSON
-output gains a `wrote` list of the paths actually written. `install-uscha.py:cmd_init`.
-Regression: smoke **T85**.
-
-## Note
-This is the first fix that came out of `uscha` measuring itself (the repo now carries its own
-`uscha.config.json`, `ACCEPTANCE.md`, and versioned `QA-LEDGER.json`). More of the same retro's
-findings — chiefly the inert `risk_profile` — are still queued.

package/uscha-kit/CHANGELOG-1.45.0.md

@@ -1,58 +0,0 @@
-# uscha-kit 1.45.0 — the risk profile stops being a decoration (2026-07-21)
-
-Release B of the field retrospective, and its deepest finding: `risk_profile` was declared in
-config and read by **nobody** in the engine — "the difference between a methodology and a
-convention." Designed front-to-back with `uscha-adr-refine` (five rounds of interrogation
-before a line of code); see `docs/adr/ADR-001` and `ADR-002`. Smoke suite: 372/372.
-
-## What it does
-
-### `risk_profile` is a named preset that modulates the flow (ADR-001)
-A profile is **not** a fixed table the engine imposes — that would contradict the provenance
-doctrine (kit 1.17.0), where the human declares what gates and the engine holds no opinion of
-its own. It is a **kit-shipped, overridable preset**: at config load, `risk_profile` expands
-into knobs the config already understands (`qa_tools_order`, `coverage_threshold`,
-`golden_required`), and any explicit `defaults` key wins per-key. The gating power it confers is
-exactly the power a human already has by declaring those knobs — a preset with provenance, not
-a new mechanism.
-
-Default table (all overridable):
-
-| Profile | `qa_tools_order` (required to converge) | `golden_required` | coverage_threshold |
-|---|---|---|---|
-| A | `[code-review]` | no | — |
-| B | `[code-review, improve]` | no | — |
-| C | `[code-review, judgment-day, improve]` | no | — |
-| D | `[code-review, judgment-day, improve]` | **yes** | 70 |
-| E | `[code-review, judgment-day, improve]` | **yes** | 80 |
-
-- An **unknown** profile fails loud (`SystemExit`) — a declared risk level is never inert
-  (INV-RISK-01, new in `CONSTITUTION.md`).
-- **No** `risk_profile` → behavior byte-identical to before (backward compatible).
-- The origin of each expanded key is tracked so a cap can label its provenance.
-
-### `golden_required` — a declarable cap for "an approved golden must exist" (ADR-002)
-A new readiness knob, declarable by **any** config (profiles D/E preset it on). When active and
-**no approved `golden-diff` gate** was ever logged for a repo, the frozen baseline is ABSENT —
-a measured fact — and that repo's readiness is capped at **49** (NOT READY, does not pass the
-human merge gate). A present-but-FAILED golden is left to the existing BLOCKER path (no
-double-cap). Provenance is three-way: `requerimiento (perfil E)` when it came from a profile,
-`requerimiento (config)` when declared directly, exposed in `cap_source`.
-
-## Scope
-Engine-only. This makes the profile **weigh in the ledger and readiness** — it kills the inert
-defect. The token-saving half (the `uscha-devloop` orchestrator actually SKIPPING sub-agents by
-profile at run time) is orchestrator behavior, not deterministic engine logic, and is a named
-follow-up. The profile does not touch `execution_policy` (routing), is global (not per-repo),
-and never auto-classifies risk — the human declares it.
-
-## Regressions
-- Smoke **T86**: profile expands the knobs with a provenance marker; explicit config wins;
-  unknown profile fails hard.
-- Smoke **T87**: `golden_required` caps readiness to 49 when the golden is absent, with the
-  right `cap_source` (profile vs config); with an approved golden or an explicit override, it
-  does not fire.
-
-## Note on the test suite
-The installer/npm smoke checks still hardcode the version string, so this bump updated those
-literals.

package/uscha-kit/CHANGELOG-1.46.0.md

@@ -1,50 +0,0 @@
-# uscha-kit 1.46.0 — a progress statusline, generic and auto-wired (2026-07-21)
-
-The progress statusline (the colored bars at the bottom of Claude Code) was a prototype living
-inside a single consumer project (ANTI-FARO), full of that project's specifics. This
-generalizes it into the kit and makes the installer wire it, so a project gets a live progress
-readout with **zero** `settings.json` editing. Smoke suite: 374/374.
-
-## Generic, config-driven scripts
-`templates/scripts/uscha_statusline.py` (the renderer) and `uscha_progress.py` (the Stop-hook
-refresher) carry **no** project-specific data. Everything that used to be hardcoded — the
-tracked repo, its label, its roadmap, its build priority — now comes from `uscha.config.json`:
-
-```jsonc
-"repos": [{
-  "name": "myproj", "path": ".", "type": "python", "label": "MY PROJECT",
-  "roadmap": [ {"name": "01 Core", "path": "src/core.py"}, ... ],
-  "build_priority": ["02 API", "01 Core"]
-}]
-```
-
-The refresher reads REAL numbers only — acceptance from `ACCEPTANCE.md` (tolerant of `AC-01`
-and `**AC-01**`), tests/coverage from the ledger snapshot, roadmap items counted as built when
-their file actually exists and is non-trivial. Truth-pass: a missing source leaves its field
-null. The renderer degrades to an **empty** line (hidden) when there is no data — no hardcoded
-fallback ever.
-
-## Auto-wired by `uscha init`
-`init` now also copies the two scripts to the project's `.claude/scripts/` and MERGES into
-`.claude/settings.json`, without clobbering:
-- `statusLine` → `python .claude/scripts/uscha_statusline.py` (added only if absent; an
-  existing DIFFERENT statusLine is reported as a conflict, never overwritten — `--force` to
-  replace);
-- a `Stop` hook → `python .claude/scripts/uscha_progress.py` (appended only if not already
-  registered; idempotent).
-
-Commands are by name with forward slashes (Windows eats backslashes in the statusLine command;
-absolute paths are brittle across machines). A clean repo initializes fully; a repo with a
-conflicting file stays per-file (kit 1.44.1) and the settings merge still runs.
-
-## Regressions
-- Smoke **T88**: the scripts are generic (config-driven, no ANTI-FARO), render with the
-  config's label, and degrade to empty with no data.
-- Smoke **T89**: `init` wires `statusLine` + Stop hook + scripts, is idempotent, and never
-  clobbers a foreign `statusLine`.
-- Smoke **T85** updated: `init` now writes the scripts + `settings.json` alongside the templates.
-
-## Note
-Migrating the ANTI-FARO consumer to the kit's version (dropping its local prototype) is the
-consumer's job, out of this repo. Making the statusline command auto-detect `python`/`py -3`
-like the npm router is a possible follow-up.