loopx-agent · Oct 17, 2026 · Oct 17, 2026
@@ -5,7 +5,15 @@
 LoopX is a local-first, runtime-neutral control plane for long-horizon agents
 and personal agent teams. The Agent host supplies models and tools; LoopX keeps
 goals, decisions, work ownership and evidence durable across sessions and hosts.
-核心目标：让 Agent 围绕长期目标持续推进，协作、恢复，并交付可验证的结果。
+The core objective is better long-horizon effectiveness and efficiency: more
+independently verifiable outcome value within comparable budgets, or comparable
+accepted outcomes with less total cost and human attention. Structured state,
+closed-loop feedback and evidence-driven replanning provide process supervision
+across sessions, recovery and collaboration; their value depends on actual
+decisions, execution and outcomes.
+核心目标：通过结构化状态、闭环反馈和基于证据的重规划过程监督，提升 Agent 与
+Agent 团队的长程效果和长程效率：在可比预算下取得更好的可验证成果，或以更低的
+完整成本与人工介入交付同等验收成果。
 
 - **Goals and work:** durable Goals, Todos, dependencies, acceptance and recovery.
 - **Governed continuation:** heartbeat, scheduling, quota and explicit stop/wait
@@ -59,6 +67,15 @@ Small diffs, fields, receipts, test counts and merged PR counts do not establish
 progress. Characterization, prerequisites, research, docs and maintenance are
 valid when they remove an evidenced gap or enable a named real next step.
 
+For changes intended to improve long-horizon work, trace the accepted goal and
+structured state through feedback, the replan decision, actual subsequent work
+and validated outcomes. Check whether new evidence changed the action or
+justified continuing it; preserve useful negative results and uncertainty.
+State survival, a graph edge, a read receipt or a replan record alone does not
+prove adoption or benefit. Compare outcomes and complete costs under comparable
+conditions using the [existing measurement contract](docs/architecture/rfcs/long-horizon-harness-benchmark-research-program-v0.md#6-measurement-model);
+do not force a direction change merely to increase replan counts.
+
 Continue through the selected slice's implementation, integration, negative
 cases and readback while authorized work remains feasible. Do not stop after
 setup, a serializer, a mock or an isolated smoke when the useful outcome is

@@ -30,6 +30,32 @@ LoopX aims to let people express, revise and accept complex goals through a loca
 
 Product success means **accepted outcomes within constraints, human attention cost and recovery capability**. Online counts, generated text, tool calls and PR counts are process signals. Models, runtimes, storage providers, IM and memory services can change; Goal/Todo/claim/lease/quota/effect and acceptance rules retain explicit, traceable single owners.
 
+The optimization objective is **long-horizon effectiveness and efficiency**:
+better independently accepted outcomes within comparable budgets, or comparable
+accepted outcomes with less total cost and human attention. Reliability,
+recovery and collaboration support that objective. LoopX pursues it through:
+
+- **Structured state:** preserve accepted goals, constraints, evidence provenance,
+  open questions, dependencies and decisions across compaction, restart and handoff.
+- **Closed-loop feedback:** connect execution, artifact validation and permitted
+  task/user feedback to the current state and next decision, retaining negative
+  results, uncertainty and evidence that supports staying the course.
+- **Evidence-driven replan process supervision:** inspect whether relevant
+  feedback reaches the responsible Agent, informs its decision and is adopted
+  in subsequent execution; then validate the resulting outcome. A justified
+  continuation is valid, and a forced pivot can discard useful work.
+
+Qualify this chain separately from storing a checkpoint, adding graph edges,
+acknowledging a read or completing a replan. Compare task-native outcomes and
+complete costs under aligned source, starting state, feedback, budget and
+observation windows. Include model/context processing, tools, control overhead,
+evaluation, recovery and human attention where measured; preserve unknown costs
+and count newly processed context after a restore. Use the existing
+[measurement model](long-horizon-harness-benchmark-research-program-v0.md#6-measurement-model)
+and [evidence-conditioned replan studies](long-horizon-harness-benchmark-research-program-v0.md#82-evidence-stall-detection-and-semantic-replan).
+These mechanisms and study hypotheses do not by themselves establish a measured
+benefit, change production defaults or grant additional execution authority.
+
 | User and work | Shortest value path | Required accepted outcome |
 | --- | --- | --- |
 | Individual developer/researcher | Connect a project→existing or managed Agent→continued work→interruption recovery | Reproducible installation/first useful action; visible work, artifacts, cost and next step without learning every RFC |

@@ -25,6 +25,26 @@ LoopX 的目标是让人用本地前端或 Lark 提出、修订和验收复杂
 
 产品成效以**在约束内交付的已验收成果、人的注意力成本和恢复能力**衡量。在线 Agent 数、生成内容量、工具调用数和 PR 数是过程信号。模型、runtime、存储 provider、IM 或 memory 服务可以替换；Goal/Todo/claim/lease/quota/effect 与验收规则必须保持明确、可追溯的唯一 owner。
 
+优化目标是**长程效果和长程效率**：在可比预算下取得更好的独立验收成果，或以更低的
+完整成本与人工介入交付同等验收成果。可靠性、恢复与协作为此服务。LoopX 通过以下
+机制追求这一目标：
+
+- **结构化状态：** 跨上下文压缩、重启与交接保留已接受目标、约束、证据来源、
+  未决问题、依赖和决策。
+- **闭环反馈：** 将实际执行、产物验证及允许的任务／用户反馈接回当前状态和下一步
+  决策，保留负结果、不确定性，以及支持继续当前方向的证据。
+- **基于证据的重规划过程监督：** 检查相关反馈是否送达负责 Agent、影响其决策并被
+  后续执行实际采用，再验证结果。有依据的继续执行是合法选择；强制换方向可能丢弃
+  有用工作。
+
+这条链路的验证独立于保存 checkpoint、增加图边、确认已读或完成一次重规划。
+在来源版本、起始状态、反馈、预算和观察窗口对齐的条件下，比较任务原生结果和完整
+成本。按实际覆盖纳入模型／上下文处理、工具、控制开销、评测、恢复与人的注意力；
+未知成本保持未知，恢复后重新处理的上下文仍计成本。复用既有
+[计量模型](long-horizon-harness-benchmark-research-program-v0.zh-CN.md#6-测量模型)
+和[证据驱动重规划研究](long-horizon-harness-benchmark-research-program-v0.zh-CN.md#82-evidence空转检测与-semantic-replan)。
+这些机制和研究假设本身不构成已测得收益，不改变生产默认配置，也不授予额外执行权限。
+
 | 用户与工作 | 最短价值路径 | 必须能验收的结果 |
 | --- | --- | --- |
 | 个人开发者/研究者 | 连接项目→一个现有或 managed Agent→连续工作→中断接续 | 安装与首次有效动作可复现；工作、产物、成本和下一步可读；不用理解全部 RFC |