prototype(tui): throwaway TUI information-architecture prototype (ticket #3)

Three structurally different variants (Panes / Pages / Ledger), switchable
live, plus a six-state scenario rotator (steady, warming up, habit changed,
stale, no baseline, paused) driving the ADR-0002 section-13 contract and the
four separate ADR-0003 section-8 service facts. Pause/resume/collect suspend
the TUI and run a polkit stand-in on the real terminal to validate tty
passthrough. Headless smoke test + SVG screenshots included.

.gitignore

@@ -1,5 +1,4 @@
 __pycache__/
-.pi/
 *.pyc
 .commandcode/
 data/fenris.pid

AGENTS.md

@@ -1,9 +0,0 @@
-## Agent skills
-
-### Issue tracker
-
-Issues are tracked in Gitea using the authenticated `tea` CLI. See `docs/agents/issue-tracker.md`.
-
-### Domain docs
-
-This is a single-context repository. See `docs/agents/domain.md`.

docs/agents/domain.md

@@ -1,26 +0,0 @@
-# Domain Docs
-
-How engineering skills should consume this repository’s domain documentation.
-
-## Layout
-
-This is a single-context repository:
-
-```text
-/
-├── CONTEXT.md
-├── docs/adr/
-└── ...
-```
-
-## Before exploring
-
-Read `CONTEXT.md` and relevant ADRs under `docs/adr/` when they exist. If they do not exist, proceed silently. Domain-modeling skills create them lazily when terminology or durable architectural decisions are resolved.
-
-## Use the glossary’s vocabulary
-
-Use terminology defined in `CONTEXT.md` consistently. If required terminology is missing or contradictory, raise it through domain modeling rather than silently inventing synonyms.
-
-## Flag ADR conflicts
-
-If proposed work contradicts an existing ADR, identify the conflict explicitly instead of silently overriding it.

docs/agents/issue-tracker.md

@@ -1,85 +0,0 @@
-# Issue tracker: Gitea
-
-Issues for this repository live in Gitea at:
-
-https://git.bongbetic.com/xavierk/Fenris/issues
-
-Use the authenticated `tea` CLI from the repository root. The configured login is `xavierk`.
-
-## General operations
-
-- List: `tea issues list`
-- Read: `tea issues <index> --comments`
-- Create: `tea issues create --title "<title>" --description "<body>"`
-- Edit: `tea issues edit <index> --title "<title>" --description "<body>"`
-- Assign: `tea issues edit <index> --add-assignees "<username>"`
-- Add labels: `tea issues edit <index> --add-labels "<labels>"`
-- Comment: `tea comments add <index> --description "<comment>"`
-- Close: `tea issues close <index>`
-- Reopen: `tea issues reopen <index>`
-
-Use `--output json` for machine-readable list and read operations. Use `tea api` when the high-level issue commands do not expose a native Gitea operation.
-
-## When a skill says “publish to the issue tracker”
-
-Create a Gitea issue in this repository. Preserve Markdown formatting in its body and apply any labels required by the invoking skill.
-
-## When a skill says “fetch the relevant ticket”
-
-Read the named issue with comments. The user may provide its URL, title, or index. In user-facing output, refer to issues by their linked titles rather than bare indices.
-
-## Wayfinding operations
-
-Wayfinder maps and decision tickets are Gitea issues.
-
-### Map and ticket grouping
-
-- A map has the label `wayfinder:map`.
-- Create one milestone named `Wayfinder: <map title>` for the effort.
-- Assign the map and all its tickets to that milestone.
-- Every ticket links its parent by name near the top: `Parent map: [<map title>](<map URL>)`.
-- Every ticket has exactly one type label: `wayfinder:research`, `wayfinder:prototype`, `wayfinder:grilling`, or `wayfinder:task`.
-
-The shared milestone and explicit parent link express the child relationship, because this Gitea version has no native parent/child issue API.
-
-### Blocking
-
-Use Gitea’s native issue-dependency relationship. To make `<blocked>` depend on `<blocker>`:
-
-```bash
-tea api -X POST \
-  repos/{owner}/{repo}/issues/<blocked>/dependencies \
-  -F index=<blocker> \
-  -f owner=xavierk \
-  -f repo=Fenris
-```
-
-List blockers:
-
-```bash
-tea api repos/{owner}/{repo}/issues/<index>/dependencies
-```
-
-Remove the relationship with the same payload and `-X DELETE`.
-
-### Frontier
-
-List open issues in the map’s milestone. Exclude:
-
-- the issue labelled `wayfinder:map`
-- assigned tickets, because assignment is the claim
-- tickets whose dependency query returns any open issue
-
-The remaining open, unassigned, unblocked tickets are the frontier. Choose the oldest first unless the user names one.
-
-### Claim
-
-Before doing any ticket work, assign it to the current `tea whoami` user. An open ticket without an assignee is unclaimed.
-
-### Resolve
-
-1. Add the answer as a resolution comment.
-2. Close the ticket.
-3. Re-fetch the map immediately before editing it.
-4. Append a linked one-line context pointer to `Decisions so far`.
-5. Create newly visible tickets, then wire dependencies in a second pass.

prototype/tui-ia/.gitignore

@@ -0,0 +1,2 @@
+.venv/
+__pycache__/

prototype/tui-ia/README.md

@@ -0,0 +1,46 @@
+# Fenris TUI information-architecture PROTOTYPE (throwaway)
+
+**This is throwaway code answering [ticket #3](https://git.bongbetic.com/xavierk/Fenris/issues/3).** It is not the redesign, reads nothing real, and never ships. Branch: `prototype/tui-information-architecture`.
+
+## Question
+
+What screen hierarchy, navigation, and action model makes Fenris's projection contract (ADR 0002 §13), the four separate service facts (ADR 0003 §8), warming-up, unexplained gaps, and changing habits understandable in a keyboard-first terminal?
+
+## Run (one command)
+
+```sh
+./run
+```
+
+(creates `.venv` and installs `textual` on first use)
+
+## What to flip through
+
+**Variants (← / →)** — three structurally different answers, not restylings:
+
+| Key | Variant | Idea |
+|-----|---------|------|
+| A | **Panes** | everything on one dense screen, btop-style; no navigation, panes are zones |
+| B | **Pages** | persistent three-fact header (lifespan · confidence · freshness) + pages 1–5 |
+| C | **Ledger** | one scrolling document in reading order, headline sentence first |
+
+**States (s)** — same variants, six shapes of the contract:
+
+1. steady · Supported (with one unexplained 3-hour gap)
+2. warming up · Limited (11 of 14 days)
+3. habit changed · Limited (regime 6 days old, scenario spread visible)
+4. stale · Supported→Limited (last collect FAILED, 61 h old)
+5. no baseline · Unavailable (wear too coarse to imply endurance)
+6. paused · Limited (period closed by deliberate disable)
+
+**Actions** — `p` pause (asks confirmation) · `r` resume (doesn't) · `c` collect now (synchronous outcome). Each suspends the TUI and runs `polkit_stub.py` on the real terminal: this validates the tty passthrough ADR 0003 requires for the polkit prompt. Results land in the tty log (variant B service page; every variant's log is the same list).
+
+## What to react to
+
+- Which variant's hierarchy matches how you think about the drive? (Mixing — "header from B, density of A" — is a valid answer and the point.)
+- Are the four service facts separable at a glance?
+- Do confidence states + contributing facts read as evidence, not as a percentage?
+- Is the pause confirmation the right amount of friction?
+- Did the polkit tty stub actually prompt in your terminal? (That's the mechanism check.)
+
+`screenshots/` holds headless captures (`smoke_test.py`) of each variant at 80×24 and 140×40.

prototype/tui-ia/polkit_stub.py

@@ -0,0 +1,29 @@
+#!/usr/bin/env python3
+"""STUB polkit-agent stand-in for the Fenris TUI prototype (throwaway).
+
+Runs attached to the real terminal while the Textual app is suspended, exactly
+where the platform polkit agent would prompt for `com.bongbetic.fenris.monitor`.
+Accepts any password; the point is validating tty passthrough, not auth.
+"""
+import getpass
+import sys
+import time
+
+op = sys.argv[1] if len(sys.argv) > 1 else "unknown"
+print("=" * 56)
+print(" polkit STUB · com.bongbetic.fenris.monitor")
+print(f" operation: {op}")
+print(" Authentication required to manage Fenris monitoring")
+print("=" * 56)
+try:
+    getpass.getpass(" password (anything works): ")
+except (EOFError, KeyboardInterrupt):
+    print("\n(cancelled — operation not performed)")
+    sys.exit(1)
+time.sleep(0.6)  # pretend systemctl + monitoring-period bookkeeping
+print(f" fenris-monitor {op}: done")
+try:
+    input(" [press Enter to return to the TUI] ")
+except EOFError:
+    pass
+sys.exit(0)

prototype/tui-ia/requirements.txt

@@ -0,0 +1 @@
+textual>=0.60

prototype/tui-ia/run

@@ -0,0 +1,14 @@
+#!/usr/bin/env bash
+# PROTOTYPE runner — throwaway, see README.md
+set -euo pipefail
+cd "$(dirname "$0")"
+if [ ! -x .venv/bin/python ]; then
+  if command -v uv >/dev/null 2>&1; then
+    uv venv -q .venv
+    uv pip install -q --python .venv/bin/python -r requirements.txt
+  else
+    python3 -m venv .venv
+    .venv/bin/pip -q install -r requirements.txt
+  fi
+fi
+exec .venv/bin/python tui_prototype.py

prototype/tui-ia/smoke_test.py

@@ -0,0 +1,67 @@
+# Headless smoke test for the prototype: drives every variant × state through
+# Textual's test pilot, exports SVG screenshots, and asserts contract strings render.
+import asyncio, inspect, os, sys
+os.environ["FENRIS_PROTOTYPE_NO_TTY"] = "1"
+
+sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
+from tui_prototype import FenrisPrototypeApp, VARIANTS
+
+OUT = os.path.join(os.path.dirname(os.path.abspath(__file__)), "screenshots")
+os.makedirs(OUT, exist_ok=True)
+
+async def snap(app, name):
+    r = app.export_screenshot()
+    if inspect.isawaitable(r):
+        r = await r
+    with open(os.path.join(OUT, name + ".svg"), "w") as f:
+        f.write(r)
+
+async def main():
+    checks = []
+    for size in [(80, 24), (140, 40)]:
+        app = FenrisPrototypeApp()
+        async with app.run_test(size=size) as pilot:
+            await pilot.pause()
+            for i, (key, _name) in enumerate(VARIANTS):
+                if i:
+                    await pilot.press("right"); await pilot.pause()
+                tag = f"v{key}_{size[0]}x{size[1]}"
+                await snap(app, tag + "_steady")
+            # rotate states on variant B
+            await pilot.press("left"); await pilot.pause()  # back to A
+            await pilot.press("right"); await pilot.pause()  # B
+            for st in ["warming", "changed", "stale", "nobaseline", "paused", "steady"]:
+                await pilot.press("s"); await pilot.pause()
+                await snap(app, f"vB_{size[0]}x{size[1]}_{st}")
+            # pages on B
+            for k in ["1", "2", "3", "4", "5"]:
+                await pilot.press(k); await pilot.pause()
+            body = app.query_one("#vB-service").content
+            checks.append(("service facts", "boot enablement" in str(body) and "freshness" in str(body)))
+            # pause flow: confirm modal, y, stub skipped headless -> state becomes paused? (skip branch)
+            await pilot.press("4"); await pilot.pause()
+            await pilot.press("p"); await pilot.pause()
+            await pilot.press("n"); await pilot.pause()  # cancel, no state change
+            head = str(app.query_one("#vB-overview").content)
+            checks.append(("confidence renders", "Projection confidence" in head))
+            # disclosure modal
+            await pilot.press("d"); await pilot.pause()
+            await pilot.press("escape"); await pilot.pause()
+    steady = FenrisPrototypeApp()
+    async with steady.run_test(size=(140, 40)) as pilot:
+        await pilot.pause()
+        doc = str(steady.query_one("#vC-doc").content)
+        checks += [
+            ("headline", "Usage-adjusted theoretical lifespan" in doc),
+            ("scenario range", "Scenario range" in doc),
+            ("wear line", "Vendor wear" in doc),
+            ("gap marker", "unexplained gap" in doc),
+        ]
+    failed = [n for n, ok in checks if not ok]
+    print("CHECKS:", "all ok" if not failed else f"FAILED: {failed}")
+    for n, ok in checks:
+        print(f"  {'ok ' if ok else 'FAIL'} {n}")
+    if failed:
+        sys.exit(1)
+
+asyncio.run(main())

prototype/tui-ia/tui_prototype.py

@@ -0,0 +1,573 @@
+# PROTOTYPE (THROWAWAY) — Fenris TUI information-architecture prototype.
+# Question (ticket #3): what screen hierarchy, navigation, and action model makes
+# the projection contract (ADR 0002 §13) and the service facts (ADR 0003 §8)
+# understandable in a keyboard-first terminal?
+# Plan: three structurally different variants (A Panes / B Pages / C Ledger),
+# switchable live with ←/→, plus a scenario rotator (s) that drives the same
+# variants through warming-up / habit-change / stale / paused / no-baseline states.
+# Data is synthetic but modeled on the real drive (Micron 2400 512GB, ~101 TB
+# written, 50 % used). Nothing here reads or writes the observation store.
+
+from __future__ import annotations
+
+import os
+import random
+import subprocess
+import sys
+from pathlib import Path
+
+from textual.app import App, ComposeResult
+from textual.binding import Binding
+from textual.containers import Horizontal, Vertical, VerticalScroll
+from textual.screen import ModalScreen
+from textual.widgets import Static
+
+STUB = Path(__file__).with_name("polkit_stub.py")
+
+# ---------------------------------------------------------------- fake data
+
+DRIVE = {
+    "model": "Micron_2400_MTFDKBA512QFM",
+    "capacity": "512 GB",
+    "percentage_used": 50,
+    "written_tb": 101.1,
+    "temp": 38,
+    "spare": 100,
+    "media_errors": 0,
+    "power_on_hours": 11126,
+    "power_cycles": 5188,
+    "unsafe_shutdowns": 142,
+}
+
+BASELINE = {"tb": 220.0, "source": "Micron 2400 datasheet (PROTOTYPE placeholder provenance)"}
+
+
+def _years(gb_per_day: float) -> str:
+    days = (BASELINE["tb"] - DRIVE["written_tb"]) * 1000.0 / gb_per_day
+    if days >= 365.25:
+        return f"~{days / 365.25:.1f} years"
+    return f"~{days:.0f} days"
+
+
+def _days_hist(n: int, rate: float, seed: int) -> list[float]:
+    rng = random.Random(seed)
+    return [max(2.0, rate + rng.gauss(0, rate * 0.18)) for _ in range(n)]
+
+
+def _sparkline(vals: list[float], width: int = 40) -> str:
+    if not vals:
+        return ""
+    mx = max(vals) or 1.0
+    blocks = " ▁▂▃▄▅▆▇█"
+    step = max(1, len(vals) // width or 1)
+    picked = vals[-width * step:][::step][-width:]
+    return "".join(blocks[min(len(blocks) - 1, int(v / mx * (len(blocks) - 1)) + (1 if v > 0 else 0))] for v in picked)
+
+
+def _habit_bar(active: float, idle: float, off: float, unknown: float, width: int = 44) -> str:
+    total = active + idle + off + unknown or 1.0
+    segs = [("active", active, "green"), ("idle", idle, "yellow"), ("off", off, "cyan"), ("?", unknown, "magenta")]
+    out = []
+    for label, v, color in segs:
+        n = max(1 if v else 0, round(v / total * width))
+        out.append((f"[{color}]{label[0] * n}[/{color}]"))
+    legend = f" active {active / total:.0%} · idle {idle / total:.0%} · powered-off {off / total:.0%} · unknown {unknown / total:.0%}"
+    return "".join(out) + "\n " + legend
+
+
+def build_scenarios() -> list[dict]:
+    """Six states of the ADR-0002 §13 contract + ADR-0003 service facts."""
+    hist_steady = _days_hist(34, 55, seed=7)
+    hist_warm = _days_hist(11, 62, seed=11)
+    hist_changed = _days_hist(35, 70, seed=3) + _days_hist(6, 130, seed=4)
+    hist_stale = _days_hist(34, 55, seed=7)
+    hist_nobase = _days_hist(26, 48, seed=5)
+    hist_paused = _days_hist(28, 51, seed=9)
+
+    def svc(enabled, timer, outcome, freshness, period):
+        return {"enabled": enabled, "timer": timer, "outcome": outcome, "freshness": freshness, "period": period}
+
+    return [
+        {
+            "key": "steady",
+            "name": "steady · Supported",
+            "days": 34,
+            "history": hist_steady,
+            "rate": 55,
+            "headline": _years(55),
+            "confidence": "Supported",
+            "facts": [
+                "34 qualifying days (≥ 14), coverage 92 %",
+                "7- and 28-day rates within a factor of 2",
+                "no single day ≥ 50 % of trailing 28-day writes",
+                "1 day with 3 unknown hours — unexplained gap inside the period",
+            ],
+            "horizons": [("last 7 days", _years(48)), ("last 28 days", _years(57))],
+            "habit": (0.34, 0.52, 0.10, 0.04),
+            "gap_days": {-9},
+            "habit_change": None,
+            "service": svc(True, True, "ok · 3 min ago (5 min cadence)", "fresh · newest sample 3 min old", "open since Aug 3 · deliberate disables: 0"),
+        },
+        {
+            "key": "warming",
+            "name": "warming up · Limited",
+            "days": 11,
+            "history": hist_warm,
+            "rate": 62,
+            "headline": _years(62),
+            "confidence": "Limited",
+            "facts": [
+                "warming up: 11 of 14 qualifying days",
+                "coverage 84 %",
+            ],
+            "horizons": [("last 7 days", _years(66))],
+            "habit": (0.38, 0.46, 0.12, 0.04),
+            "gap_days": set(),
+            "habit_change": None,
+            "service": svc(True, True, "ok · 2 min ago", "fresh · newest sample 2 min old", "open since Aug 20"),
+        },
+        {
+            "key": "changed",
+            "name": "habit changed · Limited",
+            "days": 41,
+            "history": hist_changed,
+            "rate": 130,
+            "headline": _years(130),
+            "confidence": "Limited",
+            "facts": [
+                "usage habit changed 6 days ago — new regime adopted",
+                "regime 6 days old (young — Limited evidence)",
+                "coverage 88 %",
+                "vendor wear line disagrees ×2.1 with observed write rate",
+            ],
+            "horizons": [("last 7 days", _years(128)), ("last 28 days", _years(71))],
+            "habit": (0.47, 0.41, 0.08, 0.04),
+            "gap_days": set(),
+            "habit_change": 6,
+            "service": svc(True, True, "ok · 4 min ago", "fresh · newest sample 4 min old", "open since Jul 15 · habit change noted Aug 26"),
+        },
+        {
+            "key": "stale",
+            "name": "stale · Supported→Limited",
+            "days": 34,
+            "history": hist_stale,
+            "rate": 55,
+            "headline": _years(55),
+            "confidence": "Limited",
+            "facts": [
+                "newest evidence 61 h old (missed — older than 48 h)",
+                "34 qualifying days, coverage 92 %",
+            ],
+            "horizons": [("last 7 days", _years(48)), ("last 28 days", _years(57))],
+            "habit": (0.34, 0.52, 0.10, 0.04),
+            "gap_days": {-9},
+            "habit_change": None,
+            "service": svc(True, True, "FAILED · exit 1 · 61 h ago (device busy)", "missed · newest sample 61 h old", "open since Aug 3 · gap is unknown time inside the period"),
+        },
+        {
+            "key": "nobaseline",
+            "name": "no baseline · Unavailable",
+            "days": 26,
+            "history": hist_nobase,
+            "rate": 48,
+            "headline": None,
+            "confidence": "Unavailable",
+            "facts": [
+                "no verified rated-TBW override on record",
+                "vendor wear estimate too coarse to imply endurance (1 of ≥ 2 Percentage Used increments)",
+            ],
+            "horizons": [("last 7 days", "48 GB/day (no baseline to project)"), ("last 28 days", "44 GB/day (no baseline to project)")],
+            "habit": (0.31, 0.55, 0.10, 0.04),
+            "gap_days": set(),
+            "habit_change": None,
+            "service": svc(True, True, "ok · 3 min ago", "fresh · newest sample 3 min old", "open since Aug 8"),
+        },
+        {
+            "key": "paused",
+            "name": "paused · Limited",
+            "days": 28,
+            "history": hist_paused,
+            "rate": 51,
+            "headline": _years(51),
+            "confidence": "Limited",
+            "facts": [
+                "no open monitoring period — paused 2 days ago",
+                "paused time is excluded from the usage habit by your choice",
+            ],
+            "horizons": [("last 7 days (pre-pause)", _years(53))],
+            "habit": (0.33, 0.51, 0.12, 0.04),
+            "gap_days": set(),
+            "habit_change": None,
+            "service": svc(False, False, "ok · 2 d ago (period closed by pause)", "stale · monitoring paused 2 days ago", "closed 2 days ago · end cause: deliberate disable"),
+        },
+    ]
+
+
+# ------------------------------------------------------------- renderers
+
+
+def headline_block(sc: dict) -> str:
+    if sc["headline"]:
+        return (
+            f"[bold]Usage-adjusted theoretical lifespan: [white]{sc['headline']}[/white][/bold]\n"
+            f" if current habits continue · sustained regime: {sc['days'] if not sc['habit_change'] else sc['habit_change']} days at {sc['rate']} GB/day"
+        )
+    return "[bold]Usage-adjusted theoretical lifespan: [red]no projection from this history yet[/red][/bold]\n " + "\n ".join(sc["facts"][:2])
+
+
+def confidence_block(sc: dict) -> str:
+    color = {"Supported": "green", "Limited": "yellow", "Unavailable": "red"}[sc["confidence"]]
+    lines = [f"[bold]Projection confidence: [{color}]{sc['confidence']}[/{color}][/bold]"]
+    lines += [f" · {f}" for f in sc["facts"]]
+    return "\n".join(lines)
+
+
+def horizon_block(sc: dict) -> str:
+    rows = [f" {label:<28} → [cyan]{value}[/cyan]" for label, value in sc["horizons"]]
+    return "[bold]Scenario range[/bold] (same endurance, other horizons)\n" + "\n".join(rows)
+
+
+def wear_line(sc: dict) -> str:
+    return f"Vendor wear: {DRIVE['percentage_used']} % used · {DRIVE['written_tb']} TB of {BASELINE['tb']:.0f} TB rated (context, not a second projection)"
+
+
+def disclosure_lines() -> list[str]:
+    return [
+        "Rated endurance is a vendor guarantee boundary, not a predicted failure date.",
+        "Powered-off time counts toward the projection while monitoring is enabled; deliberately paused time does not.",
+        "The scenario range is a spread of horizons, not a statistical interval.",
+        f"Baseline provenance: {BASELINE['source']}.",
+    ]
+
+
+def history_block(sc: dict, width: int = 60) -> str:
+    vals = sc["history"]
+    spark = _sparkline(vals, width)
+    marks = [" "] * len(spark)
+    if sc["habit_change"]:
+        idx = len(spark) - max(1, round(sc["habit_change"] / max(1, len(vals) // width or 1)))
+        if 0 <= idx < len(marks):
+            marks[idx] = "▲"
+    for g in sc["gap_days"]:
+        idx = len(spark) + g - 1
+        if 0 <= idx < len(marks) and marks[idx] == " ":
+            marks[idx] = "?"
+    head = f"[bold]Usage history[/bold] · {sc['days']} days · {min(vals):.0f}–{max(vals):.0f} GB/day"
+    bar = f"[green]{spark}[/green]"
+    markline = "".join(marks)
+    a, i, o, u = sc["habit"]
+    return head + "\n " + bar + "\n " + markline + "  ▲ habit change · ? unexplained gap\n " + _habit_bar(a, i, o, u)
+
+
+def health_block() -> str:
+    d = DRIVE
+    rows = [
+        f"[bold]Drive health[/bold] · {d['model']}",
+        f" temperature {d['temp']} °C · spare {d['spare']} %",
+        f" media errors {d['media_errors']} · unsafe shutdowns {d['unsafe_shutdowns']}",
+        f" power-on {d['power_on_hours']:,} h · {d['power_cycles']:,} cycles · {d['capacity']}",
+    ]
+    return "\n".join(rows)
+
+
+def service_block(sc: dict) -> str:
+    s = sc["service"]
+    en = "[green]enabled[/green]" if s["enabled"] else "[red]disabled[/red]"
+    tm = "[green]timer active[/green]" if s["timer"] else "[red]timer inactive[/red]"
+    return "\n".join([
+        "[bold]Service[/bold] (four separate facts)",
+        f" boot enablement: {en}",
+        f" runtime activity: {tm}",
+        f" last collect outcome: {s['outcome']}",
+        f" freshness: {s['freshness']}",
+        f" monitoring period: {s['period']}",
+    ])
+
+
+def settings_block() -> str:
+    return "\n".join([
+        "[bold]Settings[/bold] (read view · edit via CLI / drop-ins)",
+ " device: /dev/disk/by-id/nvme-Micron_2400_MTFDKBA512QFM_2341ABCD",
+        " cadence: every 5 min (systemd drop-in to change) · raw retention 14 d",
+        f" endurance baseline: {BASELINE['tb']:.0f} TB rated — {BASELINE['source']}",
+    ])
+
+
+def actions_legend(paused: bool) -> str:
+    resume = "[bold green]r resume[/bold green]" if paused else "r resume"
+    pause = "[bold yellow]p pause[/bold yellow]" if not paused else "p pause"
+    return f"{pause} (asks) · {resume} · c collect now · s state · ←/→ variant · d disclosures · q quit"
+
+
+VARIANTS = [
+    ("A", "Panes — one dense screen"),
+    ("B", "Pages — persistent header + tabbed body"),
+    ("C", "Ledger — scrolling narrative document"),
+]
+
+# ----------------------------------------------------------------- screens
+
+
+class ConfirmPause(ModalScreen[bool]):
+    """Pause asks for confirmation (ADR 0003 §8)."""
+
+    BINDINGS = [
+        Binding("y", "yes", "Pause"),
+        Binding("n", "no", "Cancel"),
+        Binding("escape", "no", "Cancel", show=False),
+    ]
+
+    def compose(self) -> ComposeResult:
+        yield Static(
+            "[bold]Pause monitoring?[/bold]\n\n"
+            "This closes the current monitoring period.\n"
+            "Paused time is [bold]excluded[/bold] from your usage habit\n"
+            "(powered-off time would still count).\n\n"
+            "[dim]y pause · n cancel[/dim]",
+            id="confirm-text",
+        )
+
+    def action_yes(self) -> None:
+        self.dismiss(True)
+
+    def action_no(self) -> None:
+        self.dismiss(False)
+
+
+class Disclosures(ModalScreen):
+    BINDINGS = [Binding("escape", "close", "Close"), Binding("d", "close", "Close")]
+
+    def compose(self) -> ComposeResult:
+        yield VerticalScroll(Static("\n".join(["[bold]Disclosures[/bold]"] + [f" · {d}" for d in disclosure_lines()]) + "\n\n[dim]esc to close[/dim]", id="disc-text"), id="disc-wrap")
+
+    def action_close(self) -> None:
+        self.dismiss()
+
+
+# -------------------------------------------------------------------- app
+
+
+class FenrisPrototypeApp(App):
+    TITLE = "Fenris — PROTOTYPE (throwaway)"
+    SUB_TITLE = "TUI information architecture · ticket #3"
+
+    BINDINGS = [
+        Binding("left", "prev_variant", "‹ variant", show=False),
+        Binding("right", "next_variant", "variant ›", show=False),
+        Binding("s", "cycle_state", "state", show=False),
+        Binding("p", "pause", "pause", show=False),
+        Binding("r", "resume", "resume", show=False),
+        Binding("c", "collect", "collect now", show=False),
+        Binding("d", "disclose", "disclosures", show=False),
+        Binding("q", "quit", "quit", show=False),
+        Binding("1", "page('overview')", "overview", show=False),
+        Binding("2", "page('history')", "history", show=False),
+        Binding("3", "page('drive')", "drive", show=False),
+        Binding("4", "page('service')", "service", show=False),
+        Binding("5", "page('settings')", "settings", show=False),
+        Binding("j", "scroll_down", "scroll down", show=False),
+        Binding("k", "scroll_up", "scroll up", show=False),
+    ]
+
+    CSS = """
+    #vhost { height: 1fr; }
+    #vA { layout: grid; grid-size: 2 3; grid-columns: 3fr 2fr; grid-rows: 8 1fr 7; height: 1fr; }
+    #vA-head, #vA-bottom { column-span: 2; }
+    #vA-history { overflow-y: auto; }
+    .pane { border: round #555555; padding: 0 1; }
+    #vB-head { height: 6; border-bottom: thick #555555; padding: 0 1; }
+    #vB-body { height: 1fr; padding: 0 1; }
+    #vB-legend { height: 3; }
+    .page { height: 1fr; padding: 0 1; }
+    #vC { height: 1fr; padding: 0 2; }
+    #switchbar { height: 3; dock: bottom; background: $boost; }
+    #sb-variant { width: 1fr; text-style: reverse; }
+    #sb-state { width: 1fr; }
+    #sb-actions { width: 2fr; }
+    #confirm-text { padding: 1 2; }
+    #disc-wrap { padding: 1 2; height: auto; max-height: 80%; }
+    """
+
+    def __init__(self) -> None:
+        super().__init__()
+        self.scenarios = build_scenarios()
+        self.scenario_idx = 0
+        self.variant_idx = 0
+        self.b_page = "overview"
+        self.tty_log: list[str] = []
+
+    # ---- composition
+
+    def compose(self) -> ComposeResult:
+        with Vertical(id="vhost"):
+            with Vertical(id="vA"):
+                yield Static("", id="vA-head", classes="pane")
+                yield Static("", id="vA-history", classes="pane")
+                yield Static("", id="vA-health", classes="pane")
+                yield Static("", id="vA-bottom", classes="pane")
+            with Vertical(id="vB"):
+                yield Static("", id="vB-head")
+                with Vertical(id="vB-body"):
+                    yield Static("", id="vB-overview", classes="page")
+                    yield Static("", id="vB-history", classes="page")
+                    yield Static("", id="vB-drive", classes="page")
+                    yield Static("", id="vB-service", classes="page")
+                    yield Static("", id="vB-settings", classes="page")
+                yield Static("", id="vB-legend")
+            with VerticalScroll(id="vC"):
+                yield Static("", id="vC-doc")
+            with Horizontal(id="switchbar"):
+                yield Static("", id="sb-variant")
+                yield Static("", id="sb-state")
+                yield Static("", id="sb-actions")
+
+    def on_mount(self) -> None:
+        for vid in ("vA-head", "vA-history", "vA-health", "vA-bottom"):
+            w = self.query_one(f"#{vid}", Static)
+            w.border_title = {"vA-head": "headline", "vA-history": "usage history", "vA-health": "drive", "vA-bottom": "service + actions"}[vid]
+        self.render_all()
+
+    # ---- helpers
+
+    @property
+    def sc(self) -> dict:
+        return self.scenarios[self.scenario_idx]
+
+    def w(self, vid: str) -> Static:
+        return self.query_one(f"#{vid}", Static)
+
+    def render_all(self) -> None:
+        sc = self.sc
+        paused = not sc["service"]["enabled"]
+        key, name = VARIANTS[self.variant_idx]
+
+        # variant A: everything on one dense screen
+        self.w("vA-head").update(headline_block(sc) + "\n" + confidence_block(sc))
+        self.w("vA-history").update(history_block(sc) + "\n" + horizon_block(sc))
+        self.w("vA-health").update(health_block() + "\n\n" + settings_block())
+        self.w("vA-bottom").update(service_block(sc) + "\n " + actions_legend(paused))
+
+        # variant B: persistent header, tabbed pages
+        head = "\n".join([
+            headline_block(sc),
+            confidence_block(sc).split("\n")[0] + f" · {sc['confidence']}",
+            f"freshness: {sc['service']['freshness']} · wear: {DRIVE['percentage_used']} %",
+        ])
+        self.w("vB-head").update(head)
+        self.w("vB-overview").update(confidence_block(sc) + "\n\n" + horizon_block(sc) + "\n\n" + wear_line(sc))
+        self.w("vB-history").update(history_block(sc, 70))
+        self.w("vB-drive").update(health_block() + "\n\n" + wear_line(sc))
+        self.w("vB-service").update(service_block(sc) + "\n\n " + actions_legend(paused) + "\n\n tty log:\n" + ("\n".join(self.tty_log) if self.tty_log else " (no privileged action taken yet)"))
+        self.w("vB-settings").update(settings_block() + "\n\n" + "\n".join(" · " + d for d in disclosure_lines()))
+        self.w("vB-legend").update(f"pages: 1 overview · 2 history · 3 drive · 4 service · 5 settings   [now: {self.b_page}]")
+        for p in ("overview", "history", "drive", "service", "settings"):
+            self.w(f"vB-{p}").styles.display = "block" if p == self.b_page else "none"
+
+        # variant C: one scrolling document in reading order
+        doc = "\n\n".join([
+            "[dim]═" * 70 + "[/dim]",
+            headline_block(sc),
+            confidence_block(sc),
+            horizon_block(sc),
+            wear_line(sc),
+            history_block(sc, 70),
+            health_block(),
+            service_block(sc),
+            settings_block(),
+            "[bold]Disclosures[/bold]\n" + "\n".join(" · " + d for d in disclosure_lines()),
+            "[dim]═" * 70 + "[/dim]",
+        ])
+        self.w("vC-doc").update(doc)
+
+        # variant visibility + switcher
+        for i, vid in enumerate(("vA", "vB", "vC")):
+            self.query_one(f"#{vid}").styles.display = "block" if i == self.variant_idx else "none"
+        self.w("sb-variant").update(f" ← {key} · {name} → ")
+        self.w("sb-state").update(f" state [{self.scenario_idx + 1}/{len(self.scenarios)}]: {sc['name']} (s to cycle) ")
+        self.w("sb-actions").update(" " + actions_legend(paused))
+
+    # ---- actions
+
+    def action_prev_variant(self) -> None:
+        self.variant_idx = (self.variant_idx - 1) % len(VARIANTS)
+        self.render_all()
+
+    def action_next_variant(self) -> None:
+        self.variant_idx = (self.variant_idx + 1) % len(VARIANTS)
+        self.render_all()
+
+    def action_cycle_state(self) -> None:
+        self.scenario_idx = (self.scenario_idx + 1) % len(self.scenarios)
+        self.render_all()
+
+    def action_page(self, page: str) -> None:
+        if VARIANTS[self.variant_idx][0] != "B":
+            return
+        self.b_page = page
+        self.render_all()
+
+    def action_scroll_down(self) -> None:
+        if VARIANTS[self.variant_idx][0] == "C":
+            self.query_one("#vC").scroll_down(animated=False)
+
+    def action_scroll_up(self) -> None:
+        if VARIANTS[self.variant_idx][0] == "C":
+            self.query_one("#vC").scroll_up(animated=False)
+
+    def action_disclose(self) -> None:
+        self.push_screen(Disclosures())
+
+    def action_pause(self) -> None:
+        self.push_screen(ConfirmPause(), callback=self._pause_confirmed)
+
+    def _pause_confirmed(self, confirmed: bool) -> None:
+        if not confirmed:
+            self.tty_log.append("pause: cancelled at confirmation")
+            self.render_all()
+            return
+        result = self._run_tty_stub("pause (disable --now)")
+        if "OK" in result:
+            self.scenarios[self.scenario_idx]["service"] = {
+                "enabled": False, "timer": False,
+                "outcome": "ok · period closed by pause",
+                "freshness": "paused · no collection while disabled",
+                "period": "closed just now · end cause: deliberate disable",
+            }
+        self.tty_log.append(result)
+        self.render_all()
+
+    def action_resume(self) -> None:
+        result = self._run_tty_stub("resume (enable --now)")  # no confirmation (ADR 0003 §8)
+        if "OK" in result or "skipped" in result:
+            self.scenarios[self.scenario_idx]["service"] = {
+                "enabled": True, "timer": True,
+                "outcome": "ok · resumed just now",
+                "freshness": "fresh · collection resuming",
+                "period": "open just now",
+            }
+        self.tty_log.append(result)
+        self.render_all()
+
+    def action_collect(self) -> None:
+        result = self._run_tty_stub("collect now", blocking=True)  # synchronous outcome (ADR 0003 §7)
+        self.tty_log.append(result)
+        self.render_all()
+
+    # ---- tty passthrough validation (the point of the stub)
+
+    def _run_tty_stub(self, op: str, blocking: bool = False) -> str:
+        if os.environ.get("FENRIS_PROTOTYPE_NO_TTY") or not sys.stdin.isatty():
+            return f"{op}: tty stub SKIPPED (headless run — mechanism not exercised)"
+        try:
+            with self.suspend():
+                proc = subprocess.run([sys.executable, str(STUB), op])
+            verdict = "OK — suspend + terminal passthrough works" if proc.returncode == 0 else f"FAILED (exit {proc.returncode})"
+            return f"{op}: {verdict}"
+        except Exception as exc:  # SuspendNotSupported and friends
+            return f"{op}: suspend failed: {type(exc).__name__} — this terminal may not support passthrough"
+
+
+if __name__ == "__main__":
+    app = FenrisPrototypeApp()
+    print("[fenris prototype] throwaway UI for ticket #3 — variants switch with ←/→, states with s\n")
+    app.run()