fix(minimax-m3-max, fable-low-effort): H-SECRET-4 rejected MAX_THINKING_TOKENS in env

The validator pattern-matches *TOKEN* as credential-shaped and rejected
both contexts, failing every task on the new harnesses. The env var name
is fixed by Claude Code, so deliver the thinking budget via the
settings.json env map in init.sh instead (jq-merge preserves apiKeyHelper
from the minimax/oauth layer).

Claude-Session: https://claude.ai/code/session_019tJk7P8tZzJ24PvtgoGhLN

harnesses/contexts/fable-low-effort/v1/harness.yaml

@@ -8,9 +8,12 @@ provides: []
 # Model pin + effort dial for Claude Code against the Anthropic cloud.
 # - ANTHROPIC_MODEL forces Fable 5; workflow nodes using this layer must
 #   NOT set a node-level model: (the --model flag would override this env).
-# - MAX_THINKING_TOKENS keeps the extended-thinking budget small ("low
-#   effort") — Fable's raw capability at minimal thinking is the point:
-#   frontier judgment without frontier token burn.
+# - The small thinking budget ("low effort") is set by init.sh via
+#   settings.json — the env var name trips H-SECRET-4's credential pattern
+#   so it cannot live in this env: block. Fable's raw capability at minimal
+#   thinking is the point: frontier judgment without frontier token burn.
 env:
   ANTHROPIC_MODEL: "claude-fable-5"
-  MAX_THINKING_TOKENS: "1024"
+
+scripts:
+  init: ./init.sh