docs(CLAUDE.md): codify work-order autonomy — Claude picks what's next

User request: stop being asked "which should we work on next?" Claude
should drive the work-order autonomously per the milestone discipline +
roadmap, never present a menu of options, never pause for confirmation
between phases.

Two additive edits:

1. New "Work-order autonomy — the meta-rule" paragraph in the Milestone
   discipline section, placed above the four motivation rules. Names
   it the meta-rule that makes the others actually work. The user
   retains the right to redirect; default is "Claude drives, user
   reviews." If two next steps are genuinely equivalent, state the
   choice in one sentence and start — don't ask.

2. Strengthened the "Continue to the next planned sub-step" bullet in
   "Things you should just do without asking" under How to operate.
   Adds explicit "You pick what comes next per the Milestone discipline
   section — never present the user a menu like 'should we do X or Y?'
   or ask 'what next?'. Just choose and announce the choice in one
   sentence."

Saved as durable feedback memory in feedback_work_order_autonomy.md
(outside repo, in Claude's auto-memory).

Triggered by the 2026-05-12 session where the user reported feeling
lost / jumping between things; the diagnosis was decision fatigue from
constant work-order picks. Milestones doc (ecb0f2d) gave the structure;
this commit makes Claude responsible for executing it.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

CLAUDE.md

@@ -202,7 +202,10 @@ retail. Everything else is your call.
 
 - Continue to the next planned sub-step of a phase after the previous one
   lands clean — including immediately starting work on the next phase if the
-  current one is done
+  current one is done. **You pick what comes next** per the Milestone
+  discipline section — never present the user a menu like "should we do X
+  or Y?" or ask "what next?". Just choose and announce the choice in one
+  sentence. Work-order selection is Claude's job, not the user's.
 - Pick between two roughly equivalent implementations; justify the choice in
   the commit message
 - Refactor small amounts of surrounding code when genuinely needed to land a
@@ -545,6 +548,20 @@ collision + B.4 interaction. Demo target: walk through Holtburg without
 getting stuck, open the inn door, click an NPC, pick up an item.
 Estimated 4–6 weeks from 2026-05-12.
 
+**Work-order autonomy — the meta-rule.** You decide what to work on
+next, always. **The user does NOT pick between phases, milestones, or
+"what's next?" alternatives.** The milestone discipline + the
+per-milestone phase list + the roadmap IS the work order — drive it.
+Never ask the user "want me to start X or Y?" or present a menu of
+options. If two next steps are genuinely equivalent, state which one
+you picked and why in one sentence and start — don't ask. The user
+retains the right to redirect if they think you're wrong, but the
+default is **Claude drives, user reviews**. The user finds decision
+fatigue from constant work-order choices draining — that's literally
+what triggered the milestones doc on 2026-05-12. Honoring this rule is
+the single biggest morale lever. This is the meta-rule that makes the
+four below actually work.
+
 **The four motivation-keeping rules:**
 
 1. **One active milestone at a time.** Work that isn't on the critical