perf(portfolio): share executable revision across one evidence read

docs/architecture/rfcs/typescript-control-plane-migration-v0.md

@@ -4,7 +4,7 @@
 - Supersedes / closes: none
 - Proposed by: LoopX maintainers
 - Date: 2026-08-15
-- Last revised: 2026-10-20
+- Last revised: 2026-10-21
 - Scope: an incremental, replacement-first migration of the LoopX control-plane
   core from Python to TypeScript without maintaining two semantic
   implementations
@@ -36,6 +36,14 @@ delivery. Fewer requests alone do not establish a performance gain.
 
 ### Product conversations and asynchronous inbox
 
+**Bounded Goal portfolio reads.** `goal_portfolio.py` now joins the existing
+executable-revision request scope around the complete portfolio read. CLI,
+Chat context and manager inspection callers reuse that scope across status and
+per-Agent quota reads. Nested scopes join it; the next portfolio read resolves
+the current executable again. Registry, provider and run-history reads remain
+live, with their existing scope, limits and change detection. This is bounded
+caller adoption: no new bridge, data cache, semantic owner or provider routing.
+
 **Receiver Inbox read transaction (2026-10-19).**
 `loopx/control_plane/collaboration/inbox_read.ts` now composes local source scans, pagination,
 Goal-instance admission, receiver and peer-return read receipts, and unfinished

docs/architecture/rfcs/typescript-control-plane-migration-v0.zh-CN.md

@@ -4,7 +4,7 @@
 - 替代 / 关闭：无
 - Proposed by：LoopX maintainers
 - Date：2026-08-15
-- Last revised：2026-10-20
+- Last revised：2026-10-21
 - Scope：LoopX 控制面核心从 Python 到 TypeScript 的增量、replacement-first
   迁移；不长期维护两份语义实现
 - Tracking issue：[#3225](https://github.com/loopx-project/loopx/issues/3225)
@@ -34,6 +34,13 @@ SQLite CLI 性能。仅减少请求次数不能证明性能收益。
 
 ### 产品对话与通用异步 inbox
 
+**有界 Goal portfolio 读取。** `goal_portfolio.py` 现在在完整 portfolio 读取期间
+复用既有 executable-revision 请求作用域。CLI、Chat context 和 manager inspection
+调用者在 status 与逐 Agent quota 读取之间共用该作用域。嵌套作用域加入当前请求；
+下一次 portfolio 读取重新解析当前代码。Registry、provider 与 run-history 仍实时
+读取，保留既有范围、上限和变化检查。这是有界调用者接入，不新增桥接、数据缓存、
+语义 owner 或 provider 路由。
+
 **接收方 Inbox 读取事务（2026-10-19）。**
 `loopx/control_plane/collaboration/inbox_read.ts` 现在通过既有 typed owner 组合本地源扫描、分页、
 Goal-instance 准入、接收方及 peer-return 读取回执和未完成委派上下文。

loopx/goal_portfolio.py

@@ -12,6 +12,7 @@
 from typing import Any
 
 from .agent_registry import registered_agent_ids_for_goal
+from .control_plane.effect_runtime import effect_runtime_request_scope
 from .control_plane.goals.activation import goal_activation_state
 from .control_plane.goals.goal_ref_validation import exact_goal_ref
 from .control_plane.quota.should_run import build_quota_should_run
@@ -327,6 +328,7 @@ def _read_goal(
     return row
 
 
+@effect_runtime_request_scope()
 def build_goal_portfolio(
     *,
     registry_path: Path,

tests/test_goal_portfolio.py

@@ -516,3 +516,42 @@ def read(goal, **kwargs):
     assert result['coverage']['stopped_excluded'] == 0
     assert result['goals'][0]['activation_state'] == 'unknown'
     assert result['goals'][0]['quality'] == 'conflicting'
+
+
+@pytest.mark.parametrize("provider", ["file", "sqlite"])
+def test_portfolio_resolves_code_once_per_read_and_keeps_registration_live(
+    tmp_path, monkeypatch, provider,
+):
+    from tests.control_plane.canonical_authority_fixture import (
+        isolate_sqlite_runtime, promoted_create_fixture,
+    )
+    from loopx.control_plane import effect_runtime
+
+    isolate_sqlite_runtime(tmp_path, monkeypatch)
+    path, _, state = promoted_create_fixture(
+        tmp_path, provider=provider, registered_agents=["agent-a", "agent-b"],
+    )
+    source = json.loads(path.read_text())
+    source["goals"][0].update(domain="portfolio", adapter={"kind": "generic_project_goal_v0"})
+    path.write_text(json.dumps(source))
+    state_before = state.read_bytes()
+    fingerprints = []
+    fingerprint = effect_runtime._runtime_fingerprint
+
+    def observe_code_read():
+        revision = fingerprint()
+        fingerprints.append(revision)
+        return revision
+
+    monkeypatch.setattr(effect_runtime, "_runtime_fingerprint", observe_code_read)
+    first = portfolio.build_goal_portfolio(registry_path=path)
+    assert len(fingerprints) == 1, f"resolved code {len(fingerprints)} times"
+    assert [a["agent_id"] for a in first["goals"][0]["agents"]] == ["agent-a", "agent-b"]
+    source["goals"][0]["coordination"]["registered_agents"] = ["agent-b"]
+    path.write_text(json.dumps(source))
+    second = portfolio.build_goal_portfolio(registry_path=path)
+    assert len(fingerprints) == 2
+    assert [a["agent_id"] for a in second["goals"][0]["agents"]] == ["agent-b"]
+    assert second["inventory_revision"] != first["inventory_revision"]
+    assert second["goals"][0]["progress"] == "unknown"
+    assert state.read_bytes() == state_before