fix(agents): move lean local-model mode behind experimental flag

docs/.generated/config-baseline.sha256

@@ -1,4 +1,4 @@
-8ae312a03361238f30a2454258808744b7ad6a5d079ff321eaa60f4b8af82005  config-baseline.json
-2dac6cea7a03051e10c4f1ceeb8025c23220a550b940156d6eddeb8e3ecd75f1  config-baseline.core.json
+900c26a9b060f1dfa712abfba877bd3bf9c7b0c9f2294faf9834038283ec24b6  config-baseline.json
+d956a1d60f776bba712cb04374a4f5657cad95bb088b536c5e3e4e29d4a21328  config-baseline.core.json
 ef83a06633fc001b5b2535566939186ecb49d05cd1a90b40e54cc58d3e6e44e3  config-baseline.channel.json
 5f5d4e850df6e9854a85b5d008236854ce185c707fdbb566efcf00f8c08b36e3  config-baseline.plugin.json

docs/gateway/local-models.md

@@ -165,8 +165,9 @@ Compatibility notes for stricter OpenAI-compatible backends:
   agent-runtime prompt shape, especially when tool schemas are included. If the
   backend works for tiny direct `/v1/chat/completions` calls but fails on normal
   OpenClaw agent turns, first try
-  `agents.defaults.localModelMode: "lean"` to drop heavyweight default tools
-  like `browser`, `cron`, and `message`; if that still fails, try
+  `agents.defaults.experimental.localModelLean: true` to drop heavyweight
+  default tools like `browser`, `cron`, and `message`; this is an experimental
+  flag, not a stable default-mode setting. If that still fails, try
   `models.providers.<provider>.models[].compat.supportsTools: false`.
 - If the backend still fails only on larger OpenClaw runs, the remaining issue
   is usually upstream model/server capacity or a backend bug, not OpenClaw's