refactor(structure): promote app to repo root + remove redundant workspace cruft

Standard production layout: the OKR app (was nested under AINative_OKR_CASAN5/) is now
the repository root. No more wrapper directory.

- Promote AINative_OKR_CASAN5/* -> repo root (backend/ frontend/ packages/ apps/
  .specify/ docs/ infra/ nginx/ scripts/ + configs). Merge tool dirs: .gitea (kept the
  active deploy ci.yml, added harness-ci.yml + runbooks), .claude (agents/commands +
  launch.json), .github moved up.
- Remove redundant: 00_SUBMISSION_PACKAGE, scattered root notes (FPT_CASAN_Full.md,
  tu-tuong-casan.md, casan-tu-sinh..., casan_harness_assessment.md, source-review...,
  README_CASAN5_REFINED.md), casan-next-plans/ and optimize-docs/ (competition/planning
  artifacts — roadmap + design history preserved in git log / commit messages).
- Update all references to the old layout:
  - .gitea/workflows/{ci,harness-ci}.yml, .github/workflows/{ci,deploy}.yml:
    working-directory .; drop AINative_OKR_CASAN5/ prefix; .specify/{tests,scripts}
    -> packages/casan-harness/... (.specify/logs state kept)
  - .claude/launch.json, .gitea/*-runbook.md: path prefixes
  - CLAUDE.md, README.md: docs/input -> apps/okr/domain/input
  - policy-bundle.yaml: 8 policy paths -> packages/casan-harness/...; manifest re-signed
- secrets-scan.sh: fixture excludes -> new package/domain paths.

Full gate from the new root: PASS=64 FAIL=0 SKIP=3.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

packages/casan-harness/config/loop-policy.yaml

@@ -0,0 +1,96 @@
+# CASAN Loop Budget Governor policy (Plan-17 Track 1).
+#
+# This file is a GOVERNED, security-sensitive artifact. Loosening any ceiling
+# (raising a max_*, or adding a more permissive delegation_level / project rule)
+# must go through control-plane-settings.py (approval JWT + SoD proposer!=approver
+# + versioned + rollback) — see Plan-17 17.3 / Plan-14 / Plan-16 SEC-07.
+#
+# Deny-by-default: any run whose (profile, delegation_level, project) does not
+# match a rule here falls back to the strictest built-in ceiling in loop_common.py
+# (STRICT_CEILING), NOT to "unlimited". A missing field inside a matched rule also
+# falls back to the strict value for that field.
+#
+# on_exceed: halt | escalate  (halt = stop the loop; escalate = route to the
+# HITL approvals inbox, Plan-13 §3.4).
+version: 1
+
+# Organization hard cap (Plan-17 17.19): a governed meta-loop change can never
+# loosen a budget above these values, even with a valid approval. Defense-in-depth
+# — the governor also clamps any governed override to this cap at read time, and
+# loop-metaloop.py refuses to apply a loosen proposal that exceeds it.
+org_ceiling:
+  max_steps: 200
+  max_tokens: 1000000
+  max_wall_clock_sec: 7200
+  max_cost_usd: 25.0
+  max_corrections_per_step: 6
+
+profiles:
+  # Production is secure-by-default: tight ceilings, escalate on breach so a human
+  # decides whether to grant more budget (never silently continue).
+  prod:
+    defaults:
+      max_steps: 20
+      max_tokens: 100000
+      max_wall_clock_sec: 600
+      max_cost_usd: 1.0
+      max_corrections_per_step: 2
+      on_exceed: halt
+    delegation_levels:
+      L0:
+        max_steps: 5
+        max_tokens: 20000
+        max_wall_clock_sec: 120
+        max_cost_usd: 0.10
+        max_corrections_per_step: 1
+      L1:
+        max_steps: 10
+        max_tokens: 40000
+        max_cost_usd: 0.25
+      L2:
+        max_steps: 20
+        max_tokens: 100000
+        max_cost_usd: 1.0
+      L3:
+        max_steps: 40
+        max_tokens: 200000
+        max_wall_clock_sec: 1200
+        max_cost_usd: 3.0
+        max_corrections_per_step: 3
+      L4:
+        max_steps: 80
+        max_tokens: 400000
+        max_wall_clock_sec: 2400
+        max_cost_usd: 8.0
+        max_corrections_per_step: 4
+      L5:
+        max_steps: 160
+        max_tokens: 800000
+        max_wall_clock_sec: 4800
+        max_cost_usd: 20.0
+        max_corrections_per_step: 5
+    projects:
+      okr:
+        max_steps: 30
+    # Convergence detection (Plan-17 T2): stop a loop that repeats actions or
+    # stops making forward progress. on_stall=escalate routes to the HITL inbox.
+    convergence:
+      oscillation_repeat: 3
+      thrash_window: 4
+      no_progress_window: 3
+      on_stall: escalate
+
+  # Dev may run longer while iterating, but is still bounded and still audited.
+  dev:
+    defaults:
+      max_steps: 50
+      max_tokens: 250000
+      max_wall_clock_sec: 1800
+      max_cost_usd: 5.0
+      max_corrections_per_step: 3
+      on_exceed: halt
+    convergence:
+      oscillation_repeat: 4
+      thrash_window: 6
+      no_progress_window: 5
+      on_stall: escalate