Rest days (Saturday + Sunday)¶

How the bench enforces weekly rest without scolding the operator into it. Filed in response to operator's 2026-05-18 request: "I need to in-build rest or I'll lose it."

The problem rest solves¶

The bench is a daily-driver tool aimed at sustained progress over weeks or months. The most common failure mode for a tool like this isn't user inactivity — it's monotonic guilt-driven over-drilling that collapses into burnout. The operator has named this risk explicitly.

The research is consistent:

Kellmann (2010), Preventing overtraining in athletes in high-intensity sports: scheduled recovery improves subsequent training adaptation more than continuous load. Athletes who self-impose rest days perform meaningfully better on follow-on sessions than those who don't.
Walker (2017), Why We Sleep: skill consolidation happens during sleep, not during the doing. Inadequate recovery directly degrades the value of yesterday's practice.
Raichle (2001), Default Mode Network: the brain's integrative work — synthesis, pattern-finding, insight — happens during rest, not during focused effort. The Eureka moment is well-documented to surface during mind-wandering states (shower, walk, drive).
Maslach & Leiter, The Truth About Burnout: three burnout dimensions (exhaustion, cynicism, inefficacy) all worsen monotonically with insufficient recovery. Prevention is much easier than recovery.
Ericsson, Peak: even strict deliberate-practice protocols cap at ~4 hours/day before returns diminish, with explicit recovery between sessions.

Translation to the bench: the work this week is consolidating while the user does something else this weekend. That's the framing.

Posture¶

The rest-day surface should be the minimum thing that respects the operator's autonomy AND prevents the bench from becoming a guilt machine. Specifically:

No drilling content surfaced. No next-action card, no daily slots, no streak penalties, no "you're falling behind."
Greet, not lecture. The page exists; it doesn't preach. The framing IS the intervention.
One escape hatch for the anxious user. If the operator can't accept the rest, surface their actual evidence (the week's accomplishments at the bench) — not as a "look how busy you were" but as a calibration check: "the work you already did this week is what you're recovering from."
Don't allow an opt-out in v1. The operator wrote "I need to in-build rest or I'll lose it." Designing in an "I'll drill on Saturday anyway" toggle defeats the point. If it bites in practice, revisit.

What the rest-day surface looks like¶

When / is visited on Saturday or Sunday in the user's local timezone:

Saturday.
{display_name or "Hi"}, hope today's a good one.

No drills today. The work this week is consolidating while you do
something else.

▸ Have to peek? Here's what your week looked like at the bench.

The ▸ expander reveals an evidence-based week recap:

N problems attempted — listed by title, linked to the attempt
N content items completed — readings/videos/podcasts/exercises by kind, linked to each
N reflections written — date + title (or first 80 chars), linked
Skill changes — if the user retook the assessment this week, show the per-axis delta
Credentials uploaded — if any

Each item has a link to itself, in case the operator wants to revisit one specific thing — but the framing is not "go work on these." It's "this is what you already did. The rest is part of finishing the work."

The recap is bounded — last 7 days only, no further-back rolling history. The point is acknowledging the week's actual output, not arguing about month-over-month progress.

Timezone resolution¶

Use the same browser-local approach as the existing home-page greeting (committed in 5407f28). Sat/Sun is judged against the operator's local clock, not the server's UTC. If display_tz is set in user settings, that wins; else the browser's Intl.DateTimeFormat().resolvedOptions().timeZone.

What does NOT change on rest days¶

Profile dropdown, navigation, all other pages still work. The user can visit /aim, /library, /feedback/inbox, etc. We're only changing the home page's content, not blocking the rest of the app.
No streak penalty. Whatever streak-tracking the bench grows in the future treats Sat/Sun as "completed by design" rather than as gaps. (No streak tracker exists today; this is a future-proofing constraint.)
Reflect remains accessible. Weekly Hamming-mode review is a natural Sunday activity for some operators. If they want to open /reflect voluntarily, fine. We just don't push them there.
Active background tasks (auto-archive of closed GH issues, etc.) keep running. The rest day is for the user, not the bench.

Implementation shape¶

Route logic¶

In app/main.py:home():

@app.get("/", response_class=HTMLResponse)
def home(request: Request):
    # ... existing auth + settings load ...

    # If today is Sat or Sun in the user's local timezone, route to
    # the rest-day variant. Day-of-week is computed client-side
    # (matches the existing greeting time logic from 5407f28) by
    # passing the date string and reading it back, OR computed
    # server-side from user_settings.display_tz.
    is_rest_day = _is_weekend_for_user(request)
    if is_rest_day:
        return _render_rest_home(request, uid)

    # ... existing weekday logic ...

_is_weekend_for_user(request): - Pull display_tz from user_settings if set, fall back to Intl-style header detection or pure UTC weekend (UTC fallback is acceptable — the worst case is one user gets the rest day starting a few hours early or late) - Compute weekday() in {5, 6} (Sat = 5, Sun = 6 in Python)

_render_rest_home(request, uid): - Query last 7 days of attempts/content_progress/reflections/writings/credentials for uid - Compute simple counts + the listed items - Render templates/home_rest.html with the recap context

Template¶

New templates/home_rest.html. Spare. Centered card. Mostly text. The <details> expander pattern from the dropdown menu (proven) is fine for the week-recap reveal.

Tests¶

Mon-Fri in user's TZ → existing home renders.
Saturday → rest-day page renders, no daily_slots / next_action in the context.
Sunday → same.
Saturday with no activity this week → expander shows "Nothing logged this week — that's fine too."
Recap surfaces correct items from the last 7 days (mock the repos).
display_tz set to a timezone where it's Friday locally on Saturday UTC → returns the weekday page (or vice versa).

Decisions baked in¶

Decision	Choice	Why
Days	Saturday + Sunday	Operator specified. Easy to add a third "rest day" later via config if useful.
Opt-out	None in v1	Operator's "I need to in-build rest or I'll lose it" is the explicit ask for forced rest.
Recap window	Last 7 days	Acknowledge the week; don't get lost in longer history.
Recap content	Read-only listing	Each item links to itself; no rebuild-of-Today UI.
Streak interaction	No streak tracker exists today; future-proof by treating rest days as "complete."	Don't introduce guilt at the rest-day layer.
Visual style	Spare, centered card. One short paragraph.	The minimum that respects the operator's autonomy.

What this is NOT¶

Not a guilt trip. The recap section is opt-in via expander, not surfaced by default.
Not an absolute lock. Operator can still navigate to /library, /aim, /reflect on a rest day if they choose.
Not a streak break. Future streak tracking (if any) treats Sat/Sun as "completed by design."

Open considerations¶

Travel / timezone hopping: if the operator's display_tz is wrong (e.g., they fly to Asia and forget to update settings), the rest day will fire on the wrong calendar day. Tolerable; the worst case is one day shifted.
Operator gets a productive Saturday despite the rest framing: they can visit /library and read content directly. That's fine — choosing to rest is the intervention, not enforced rest. The rest-day home is a default, not a wall.
Holiday handling: not addressed in v1. If the user takes a Wednesday off, the bench doesn't know. Could add a manual "today is a rest day" toggle later.

References¶

Kellmann, M. (2010). Preventing overtraining in athletes in high-intensity sports and stress/recovery monitoring. Scandinavian Journal of Medicine & Science in Sports.
Walker, M. (2017). Why We Sleep: Unlocking the Power of Sleep and Dreams. Scribner.
Raichle, M. E., et al. (2001). A default mode of brain function. Proceedings of the National Academy of Sciences.
Smallwood, J., & Schooler, J. W. (2015). The science of mind wandering: Empirically navigating the stream of consciousness. Annual Review of Psychology.
Maslach, C., & Leiter, M. P. (1997). The Truth About Burnout: How Organizations Cause Personal Stress and What to Do About It. Jossey-Bass.
Ericsson, A., & Pool, R. (2016). Peak: Secrets from the New Science of Expertise. Houghton Mifflin Harcourt.