feat(antigravity): integrate App host and bounded native recovery

docs/architecture/rfcs/manager-runtime-profile-v0.md

@@ -62,6 +62,11 @@ These defaults do not change ordinary Goal Chat models or grant other Agents too
   steps to be done, matching native result recovery's completion requirement.
   Transient idle does not extend the existing no-progress or hard deadline. Permission gates,
   native errors, unverified stop reasons and another input retain their existing refusal paths.
+  Read-only polling transport loss does not resend the input or reset its idle/hard deadline.
+  Native retry/error steps remain owned by the running cascade; only a settled terminal error
+  fails the adapter, with its structured native cause retained. Host heartbeat and bounded
+  Stop-hook recovery use the separate [Antigravity host procedure](../../../skills/loopx-project/references/heartbeat-hosts.md#antigravity-app).
+  These source boundaries do not establish upstream network repair or sustained long-run reliability.
 - Antigravity manager inputs place the complete original user message before the frozen Core
   evidence, encoding that message as a JSON string so whitespace and quoted transport markers
   remain exact. Original-result recovery recognizes both this framing and historical

docs/architecture/rfcs/manager-runtime-profile-v0.zh-CN.md

@@ -50,6 +50,11 @@ LoopX 管家最初只有受限的规划会话：它可以读取 LoopX 提供的
   LoopX 继续观察原调用，不重发输入；最终答案须带正常停止标记，且本次覆盖的全部步骤均
   已完成才能收尾，与原生结果恢复的完成要求一致。短暂 idle 不延长既有无进展或硬截止
   时间；权限等待、原生错误、未核验的停止原因及另一个输入仍沿既有拒绝路径处理。
+  只读轮询的传输中断不会重发输入，也不会重置 idle 或 hard deadline。运行中的原生
+  cascade 继续拥有重试与错误步骤；接入层仅在原生终态错误后失败，并保留结构化原始
+  原因。宿主心跳与有界 Stop-hook 恢复遵循独立的
+  [Antigravity 宿主流程](../../../skills/loopx-project/references/heartbeat-hosts.md#antigravity-app)。
+  这些源码边界不证明上游网络已修复，也不证明持续长程可靠性。
 - Antigravity 管家输入先呈现完整用户原话，再附上冻结的 Core 证据；用户原话按 JSON
   字符串编码，保留空白和被引用的传输标记。原始结果恢复同时识别新格式与历史上的
   证据在前格式，沿用同一来源、输入时间、prompt hash 和原生终止检查。

examples/personal-workspace-browser/replan-cadence.mjs

@@ -83,7 +83,7 @@ export const replanCadenceScenario = {
       const count = page.getByLabel("两次复核间的数量", { exact: true });
       const strategy = page.getByLabel(/^步幅策略/);
       await waitForInputValue(count, "6");
-      if (await strategy.inputValue() !== "off") throw new Error("Stride must default off");
+      if (await strategy.inputValue() !== "evidence_conditioned_v1") throw new Error("Stride must default to evidence_conditioned_v1");
       if (await unit.inputValue() !== "effective_turns") throw new Error("Product default is not settled work Turns");
       const horizonGoal = page.getByLabel(/^在规划视野中显示 Goal/);
       const advanced = page.getByText("高级设置", {exact: true});
@@ -302,7 +302,7 @@ export const replanCadenceScenario = {
       if (inherited.source !== "capability_default" || inherited.configuration.count_unit !== "effective_turns" || inherited.configuration.count !== 6) throw new Error("Removing device policy did not restore product default");
       await reopenGoalCadence();
       await waitForInputValue(englishCount, "6");
-      if (await englishUnit.inputValue() !== "effective_turns" || await englishStrategy.inputValue() !== "off") throw new Error("Goal editor did not restore product defaults");
+      if (await englishUnit.inputValue() !== "effective_turns" || await englishStrategy.inputValue() !== "evidence_conditioned_v1") throw new Error("Goal editor did not restore product defaults");
       await englishCount.fill("5");
       await page.getByRole("button", { name: "Preview changes", exact: true }).click();
       const failedReadback = holdGoalReadback();

loopx/agent_onboarding.py

@@ -64,6 +64,8 @@ def _surface_install_command(agent_type: str, cli_bin: str, project: str) -> str
         return f"{host_command(cli_bin)} slash-commands --install --surface zcode"
     if agent_type == "agy":
         return f"{host_command(cli_bin)} slash-commands --install --surface agy"
+    if agent_type == "antigravity-app":
+        return f"{host_command(cli_bin)} slash-commands --install --surface antigravity-app"
     if agent_type == "kiro-cli":
         return f"{host_command(cli_bin)} slash-commands --install --surface kiro-cli"
     if agent_type == "pi":
@@ -297,6 +299,7 @@ def _bootstrap_pack_command(
         "cursor-agent": "cursor-agent",
         "zcode": "zcode",
         "agy": "agy",
+        "antigravity-app": "antigravity-app",
         "kiro-cli": "kiro-cli",
         "deepseek-harness": "deepseek-harness",
         "deepseek-harness-native": "deepseek-harness-native",
@@ -373,6 +376,15 @@ def _start_instruction(agent_type: str) -> str:
             "wake Prompt; recurring via MaxIterations) when quota allows more "
             "work."
         )
+    if agent_type == "antigravity-app":
+        return (
+            "Invoke the LoopX skill from `~/.gemini/config/skills` via `/loopx <task>`; "
+            "after todo writeback, start every turn with `quota should-run`, suspend "
+            "directly via the native `schedule(DurationSeconds=..., Prompt=...)` tool "
+            "when `waiting_external` or CI checks are pending (instead of polling), "
+            "and optionally configure `~/.gemini/config/sidecar.json` for background "
+            "periodic heartbeats."
+        )
     if agent_type == "kiro-cli":
         return (
             f"Run `/loopx <task>` (the LoopX skill installed in "