From 4cc1cb1bd56b5d999f5ac1c322555d45cbd70220 Mon Sep 17 00:00:00 2001 From: LoopX Agent <337587101+loopx-agent@users.noreply.github.com> Date: Mon, 2 Nov 2026 02:44:59 +0530 Subject: [PATCH] docs(blog): connect context delivery with released behavior and recovery semantics Signed-off-by: LoopX Agent <337587101+loopx-agent@users.noreply.github.com> --- .../public/blog/context-delivery/index.html | 17 ++++++++++++----- .../edgebench-feedback-and-memory/index.html | 5 +++-- .../index.html | 5 +++++ .../public/blog/zh/context-delivery/index.html | 17 ++++++++++++----- .../zh/edgebench-feedback-and-memory/index.html | 5 +++-- .../index.html | 5 +++++ 6 files changed, 40 insertions(+), 14 deletions(-) diff --git a/apps/presentation/site/public/blog/context-delivery/index.html b/apps/presentation/site/public/blog/context-delivery/index.html index 7f5396eb8b..1ba27c3367 100644 --- a/apps/presentation/site/public/blog/context-delivery/index.html +++ b/apps/presentation/site/public/blog/context-delivery/index.html @@ -9,7 +9,7 @@
← All articles

Context delivery

An Agent needs a way to look things up, too

Follow one voice task through progressive reads, JSON Pointers, exact sharing, compaction-cycle deltas, and complete recovery.

Long-running work accumulates material: the original goal, later constraints, attempted routes, failed experiments and current progress. Sending everything again on each continuation repeats more and more content; keeping only a summary can make the original basis of a decision difficult to recover.

LoopX's context delivery separates these jobs: preserve complete sources, read the facts needed for the current decision, expand supporting material around questions, and reuse unchanged structures exactly within a valid compaction cycle.

-

This article checks the native Codex path at 73cf83ed, a 2.2 candidate source. Consult Releases for publication status. In that implementation, supported POSIX Codex App/SSH sessions and new benchmark heartbeat trials default to context delivery; saved off remains effective and explicit off is available. When enabled, complete content is saved first and the Agent receives an entrypoint for current facts.

+

This article describes the native Codex path in released LoopX 2.2.0. Supported POSIX Codex App/SSH sessions and new benchmark heartbeat trials default to context delivery; saved off remains effective and explicit off is available. When enabled, complete content is saved first and the Agent receives a reading entrypoint for current facts. See the 2.2.0 release notes for publication and validation scope. The synthetic measurements below retain their original source and experimental conditions.

Think of a capture as a complete book. An index locates its chapters; a JSON Pointer addresses a place inside it. Progressive reading determines how deeply to open the material now. Differential delivery determines what can be reused on the next continuation.

Follow one task: give an assistant voice input

Suppose an assistant needs to understand a voice note and reply in its original chat. Its complete material contains the goal, provider constraints, privacy requirements, acceptance and a long discussion of alternatives. This is a teaching example; the example program below generates the complete data in an isolated directory.

@@ -174,7 +174,12 @@

On the next continuation, deliver what changed

When enabled, the default output is a reading entrypoint

Connect these operations back to the currently supported native Codex quota path:

-

The enabled delivery path

  1. Complete source: projection.json preserves current facts.
  2. Necessary view: required-context.json keeps complete requirements and points back to supporting originals.
  3. Cycle delivery: verify baseline and scope; choose full or delta.
  4. Current reader: first read admission, stops, the original task and accumulated requirements; include an index and short overview when they fit. Follow Pointers for supporting detail.
Saved files are verifiable sources. Model input still depends on the reads actually completed.
+

One capture, from saved source to reading

decision-<id>/
+├─ projection.json        source
+├─ required-context.json  facts
+├─ delivery.json          full/delta
+└─ *.reader.json          reader
+
Save the complete source, prepare the necessary view, then choose the cycle delivery. The model receives a small reference and reads the facts its current decision needs; supporting originals remain reachable. This is a layout illustration. Use the exact files and commands returned by the current invocation.

One useful property is that even with mode: "delta", the Agent can select facts directly through current recovery. It does not first need to read an old baseline or replay every delta. Delta's path may address delivery.json; recovery and select_command address the complete current view. context_details then connects supporting material to the original capture.

Native navigation also stores complete capture/selection digests in immutable private reader references. The Agent copies the returned reference; the CLI still verifies the complete digests, selection and cursor on every page. SHA256 itself is not shortened. A first_read_command starts a selection; it does not establish that all required sources were read.

# Copy the original recovery.reference_file from current returned navigation.
@@ -183,7 +188,9 @@ 

When enabled, the default output is a reading entrypoint

A continuation reference pins its selection: follow its next_command to EOF and return to the original recovery reference to change Pointers. Very small budgets may retain the explicit file/digest interface. The teaching measurements above use that compatible interface and retain their original conditions.

-

Saved-file continuity and what the model currently retains are separate facts. Verifiable files do not establish that the model remembers every original passage.

+

What was saved, and what the model has read

+

The delta codec answers: can the verified old file reconstruct the exact current file? The native captured_file path verifies the complete saved baseline, version and scope. It does not require the model to have read the old file, reached a selected page’s EOF or sent an ACK. Each baseline is itself a complete version, so the model need not replay a chain of deltas first.

+

Progressive reading answers a different question: which facts does this decision need? In the voice illustration, suppose the next continuation adds sample 3 while the privacy constraint remains unchanged. Consider a missed-read counterexample: the Agent skipped the required privacy constraint, or can no longer establish it after compaction. Before sending audio externally, it must recover /goal/constraints through the current entrypoint and honor the existing authorization boundary. It can read that fact directly, without first draining the old baseline. A saved file or an unchanged field does not replace the necessary read.

Recover missing context while preserving the original work

After host compaction, a subsequent work Turn builds a new complete baseline. Original goals, requirements and sources remain available for necessary reads.

If compaction, missing pages or truncated output interrupt an unfinished Turn, recover missing facts from its original capture while retaining task, effect and settlement identity. If a voice result was already sent but its receipt was lost, recovering context must not become sending the result again.

@@ -199,7 +206,7 @@

What the measurements establish

The example separates three mechanisms: avoid repeatedly moving history not currently needed, combine reads that belong together, and avoid retransmitting unchanged verified structures. Complete sources, requirements and recovery connect them.

Reader figures include actual CLI page metadata; local path length affects a small amount of framing. Projection figures compare capture bodies; the cycle table compares codec output. The reproduction script checks these distinct measurements and exact reconstruction.

Real long-running tasks must also count necessary later reads, recovery, cache hits, model input/output, tool calls, waiting and human attention. Effectiveness depends on constraints understood, decisions made and accepted outcomes. This synthetic example measures the mechanical path; real model benefit still needs comparable-task evidence.

-

Implemented scope includes complete sources, file entrypoints, batch selection, index depth, applicable acceptance layering, exact sharing, cycle deltas and recovery. Broader decision-stage candidate, preference and Explore delivery and C1's universal input convergence remain unfinished. Supported Codex sessions default on without a saved preference. Disable in Settings → This machine → Capabilities → Differential context → Codex; saved off survives upgrade. A command or new benchmark trial can use --context-delivery off. Trials do not inherit the machine switch, and existing frozen trials are not rewritten. Saving settings, exact hook/configuration setup and actual adoption remain separate checks; unavailable hooks retain full delivery. Ordinary CLI, unsupported hosts and non-heartbeat benchmarks retain their existing full paths. Final 2.2 release qualification follows the exact release version.

+

2.2.0 ships complete sources, file entrypoints, batch selection, index depth, applicable acceptance layering, exact sharing, cycle deltas and recovery. Broader decision-stage candidate, preference and Explore delivery and consistency across all entrypoints remain unfinished. Disable in Settings → This machine → Capabilities → Differential context → Codex; saved off survives upgrade. A command or new benchmark trial can use --context-delivery off. Trials do not inherit the machine switch, and existing frozen trials are not rewritten. Under the default policy, ordinary CLI, unsupported hosts and non-heartbeat benchmarks retain their full paths; an explicit unsupported on request is rejected. Unavailable hooks use full delivery. Saving configuration, actual installation and model adoption remain separate checks.

Read the same material yourself

example.py generates this complete synthetic voice task and invokes the existing reader, progressive projection and delta owners. It checks selected values, directory addresses, all ten requirements, shared references and complete recovery. It registers no Goal, admits or settles no Turn, and calls no model or speech service.

Prepare the original measurement checkout, a4a0fad889d91cba0d945ce658788bdb76ba351e, and its compatible dependencies. Run from that checkout’s root, substituting the downloaded program path. The output directory must be empty.

@@ -208,6 +215,6 @@

Read the same material yourself

Generated capture.json retains all 80 history entries. projection.json and required-context.json hold the complete/necessary views; cycle-base.json and cycle-delta.json hold cycle encodings; manifest.json pins files and digests. Production reads use currently returned paths, never copied teaching fields or another capture’s array indexes.

The original measurement record pins this article’s figures. Running on other source can check its mechanisms but cannot rewrite the original observation. Local path length also affects page framing bytes. Understanding this example requires no new research run.

-

Sources: complete-delivery contract, progressive projection, selection/index, delta/recovery. The example ran on a4a0fad; those three core owners are byte-identical in 0e78a19. Native final delivery also accounts for references and later reads, beyond codec output.

+

Sources: complete-delivery contract, progressive projection, selection/index, delta/recovery. The example ran on a4a0fad; its three measured core owners are byte-identical in 0e78a19. The implementation links above are pinned to released v2.2.0; the old numbers are not a post-release rerun. Native final delivery also accounts for references and later reads, beyond codec output.

diff --git a/apps/presentation/site/public/blog/edgebench-feedback-and-memory/index.html b/apps/presentation/site/public/blog/edgebench-feedback-and-memory/index.html index da79912e08..7e667dbbbd 100644 --- a/apps/presentation/site/public/blog/edgebench-feedback-and-memory/index.html +++ b/apps/presentation/site/public/blog/edgebench-feedback-and-memory/index.html @@ -57,7 +57,8 @@

When should a long, unfinished task review its direction

7. What does the control plane itself cost?

Another faulty assumption is that a smaller entry packet means a cheaper agent episode. Moving the goal, facts and writer instructions behind a second command shortens the first view, but the model still needs to read them. Losing context adds reconstruction of the same source. Measuring only entry delivery moves cost outside the table.

#7269 routes the original capture through a shared context reader: complete goals, facts and feedback first; writer format and settlement steps afterward. Review and writeback are two views of one immutable decision, rather than a second admission or two permission owners. Source regressions compare the goal, planning horizon, evidence references, complete writeback contract and action signature. A smaller entry must still reach every obligation.

-

Differencing adds a prerequisite: the receiver must retain the exact baseline. Within the same native session and context epoch, the current Hook path treats normal tool delivery after complete paging as an engineering assumption of continuity. Session changes, observed compaction, missing pages and known delivery failures invalidate it. Recovery then reads the original complete capture; a fresh decision from another guard cannot recover the old one. Missed hooks and unseen downstream truncation remain limits. A read receipt is not independent proof of model retention.

+

The early delta path depended on the receiver retaining the exact baseline. In the historical records below, the native Hook path treated normal tool delivery after complete paging, within the same session and context epoch, as an engineering assumption of continuity. Session changes, observed compaction, missing pages and known delivery failures invalidated it. Recovery used the original complete capture; a fresh decision from another guard could not recover the old one. Missed hooks and unseen downstream truncation remained limits. A read receipt was not independent proof of model retention.

+

The released 2.2.0 native file path separately verifies complete saved baselines. Exact file reconstruction and whether the model previously read a file are different facts. Even a delta allows the Agent to select necessary current facts directly through its recovery reference, without first draining the old baseline. This changes a mechanical delivery prerequisite; it does not retroactively qualify the historical experiments below. The voice example explains both kinds of continuity at one concrete decision.

One historical offline comparison in that PR totals every required stage. The table fixes base 59464f974 and tested head a006a2bea1, across four retained observations. It is not a remeasurement of the final merged head.

@@ -110,6 +111,6 @@

Measure structural input separately

The context study recorded on October 28 starts Grid, Schema and Ad Placement from their original tasks with Sol/max and an 18-hour upper limit per case. Historical Official or older LoopX runs provide descriptive references where usable. Their model/effort, source, scorer, opportunity and clock differences remain visible; they are not strict matched controls. A new baseline is warranted where the old material cannot answer the relevant question. Attempts with different source versions, continuity bases or transports are analyzed separately, rather than pooled as one treatment.

First verify that actual delivery changes: the initial full read, a subsequent delta with its continuity basis stated, complete fallback after known baseline loss, and a full review-to-work interaction. Then compare useful progress, retained artifacts and the complete usage vector. Requested flags, receipts and smaller local packets cannot substitute for that evidence. If the feature fails its intended behavior, stop the affected attempt, preserve its costs and history, repair it and use a new attempt identity.

The first Grid and Ad interactions on the new native Hook path have completed full paging, one delta and subsequent task work. Within each same capture, the delta payload was 19.53% and 19.48% smaller than its corresponding full packet. These are single-read payload comparisons: they exclude page envelopes, subsequent model input, hooks, recovery and review costs. They do not establish lower total cost. A complete review-to-native-outcome cycle, sustained benefit and failure recovery remain to be verified.

Our effectiveness target is useful progress after many hours and reliable retention of it. Our efficiency target is comparable accepted value with less complete cost and human attention. The next result must address both. Current runs retain the original deadline, and failed or replaced attempts retain their costs and history.

-

This remains the October 28 research checkpoint. Later reader, default-policy or codec merges do not automatically pass its unresolved checks. The mechanism article’s synthetic byte measurements and this section’s model trajectories are different evidence layers, not a combined savings result.

Numerical evidence retains each stated historical model, version, window and validity boundary. Public implementation: LoopX. Earlier detailed research records remain in the previous article revision. Raw trajectories, hidden evaluator material, runtime credentials and private operating context are not published here.

+

This remains the October 28 research checkpoint. The 2.2.0 release and default-on policy do not automatically pass its unresolved checks. The mechanism article’s synthetic byte measurements, release-path installation/recovery checks, and this section’s real model trajectories answer questions about mechanical delivery, product usability and research effectiveness respectively. Their versions, windows and metrics differ; they are not a combined savings result.

Numerical evidence retains each stated historical model, version, window and validity boundary. Public implementation: LoopX. Earlier detailed research records remain in the previous article revision. Raw trajectories, hidden evaluator material, runtime credentials and private operating context are not published here.

diff --git a/apps/presentation/site/public/blog/typescript-control-plane-refactor/index.html b/apps/presentation/site/public/blog/typescript-control-plane-refactor/index.html index af65cd6eed..4e84dd3cfe 100644 --- a/apps/presentation/site/public/blog/typescript-control-plane-refactor/index.html +++ b/apps/presentation/site/public/blog/typescript-control-plane-refactor/index.html @@ -250,6 +250,11 @@

10. CLI output budgets: protect decision semantics before co

Natural language requires clause-level review: obligations, ordering, evidence provenance, authority qualifiers, scope, continuation, and stop conditions. Removing a phrase such as “only when the original proof is valid” can change behavior even when every field remains.

One bounded example removed JSON indentation and separator whitespace. On the same historical fixture of 36 Todos and 12 runs, output fell from 48,418 to 39,726 characters, about 18%, with the complete decoded payload unchanged. This establishes a presentation-size change. Tokens, cost, and agent outcomes need separate measurements.

The budget protects an interface the consumer can use. Valuable guidance may justify increasing it with matched evidence. Dropping decision conditions to make a size test green damages the interface.

+

A recoverable file is not the model’s memory

+

The same recovery distinction appears in context delivery: an old receipt cannot renew a lease, and a complete baseline on disk cannot prove that the model has read and understood its constraints.

+

The native file-delta path in 2.2.0 separates those facts. The codec verifies complete files, versions and scope within a valid cycle and reconstructs the current structure exactly. The Agent progressively reads the current full view for the intent, requirements and evidence its decision needs. Even a delta permits direct selection of current facts without first draining the old baseline. Compaction during unfinished work requires recovering missing facts while preserving the original task and settlement identity.

+

This makes file reconstruction and model knowledge separately inspectable. The worked voice-task example explains JSON Pointers and the default file path. A working file/reader mechanism still needs evaluation against actual research outcomes.

+

11. Final retirement: preserve credit and account for the deletions

Late-September work had already removed an unused authority prototype and the experimental Todo event overlay, dual-writer, and backfill. Old sources containing data received explicit preservation or refusal handling. This did not remove all event logging or the new authority's atomic events and receipts.

October work further isolated ordinary Markdown writers and validated real subsequent writes without the old modules. Concentrated deletion became possible after consumer routes and replacement paths were connected.

diff --git a/apps/presentation/site/public/blog/zh/context-delivery/index.html b/apps/presentation/site/public/blog/zh/context-delivery/index.html index df2abac8a4..24bf2a4c46 100644 --- a/apps/presentation/site/public/blog/zh/context-delivery/index.html +++ b/apps/presentation/site/public/blog/zh/context-delivery/index.html @@ -9,7 +9,7 @@
← 全部文章

Context delivery

Agent 读上下文,也需要一套“查书”的方法

从一个语音任务理解渐进式读取、JSON Pointer、精确共享、compaction 周期内的结构化差分,以及失效后的完整恢复。

Agent 做长任务,资料会越来越多:原目标、后来追加的约束、试过的方案、失败的实验、当前做到哪里。每次继续都把整份材料再塞一遍,重复会不断累积;只留一段摘要,又可能找不回决定所依赖的原话。

LoopX 的 context delivery 把这件事拆开做:完整资料保存下来,当前决定需要的事实先读,支持材料沿问题展开;同一个有效 compaction 周期内,再精确复用没有变化的结构。

-

本文核对源码 73cf83ed 的 Codex 原生路径;它是 2.2 候选源码,最终发布状态以 Releases 为准。该实现的默认策略是:受支持的 POSIX Codex App/SSH 会话与新的 benchmark heartbeat 开启 context delivery;已有关闭设置继续生效,也可以显式关闭。开启后完整内容先保存,Agent 收到当前读取入口。

+

本文以已发布的 LoopX 2.2.0 Codex 原生路径为准。受支持的 POSIX Codex App/SSH 会话与新的 benchmark heartbeat 默认开启 context delivery;已有关闭设置继续生效,也可以显式关闭。开启后完整内容先保存,Agent 收到当前读取入口。具体发布与验证范围见 2.2.0 版本说明;下文的合成测量仍保留原始实验版本与条件。

可以把它想成查一本厚书:capture 是完整保存的那本书,目录告诉我们章节在哪,JSON Pointer 是书内地址。渐进式读取决定这一刻翻到多深;差分决定下一轮哪些内容可以沿用。

用“管家读语音”串起一次工作

假设要让管家理解语音消息,并在原对话返回结果。完整资料里保留了目标、服务约束、隐私要求、验收,以及很长的方案讨论。下面是教学示意,后文的示例程序会生成完整数据,可在隔离目录复现。

@@ -174,7 +174,12 @@

下一轮继续,交付这次改变了什么

开启后,默认交付的是一个读取入口

把前面的动作接回产品,当前支持的 Codex 原生 quota 路径是:

-

Context delivery 开启后的路径

  1. 完整来源:projection.json 保存当前事实。
  2. 必要视图:required-context.json 保留完整要求,支持材料指回原文。
  3. 周期交付:核对基线与 scope,选择 full 或 delta。
  4. 当前读取入口:首读当前准入、停止条件、原任务与累计要求;目录和短概览在放得下时附上。需要解释时,沿 Pointer 展开。
文件是可核验的来源;进入模型的内容,还取决于实际完成了哪些读取。
+

同一次捕获,从保存到读取

decision-<id>/
+├─ projection.json        原始捕获
+├─ required-context.json  必要视图
+├─ delivery.json          full/delta
+└─ *.reader.json          读取引用
+
先保存完整来源,再生成必要视图、选择周期交付。模型收到小引用,读取当前问题需要的事实;支持材料仍可追溯。这里是路径组织示意,实际文件与读取命令以本次返回为准。

这里有一个很有用的设计:即使返回 mode: "delta",Agent 也能从当前 recovery 直接选读事实。 它不必为了开始判断,先通读旧基线或重放全部差分。差分的 path 可以指向 delivery.json,而 recovery、select_command 指向当前完整视图;context_details 再把支持材料接回原始捕获。

原生导航还把完整 capture/selection 摘要保存在不可变的私有 reader 引用中。Agent 复制返回的引用读内容,CLI 仍在每页校验完整摘要、选择与游标;并没有把 SHA256 截短。first_read_command 是起始选读,不代表全部必读来源已经读完。

# 从当前返回的 recovery.reference_file 复制原 recovery 引用。
@@ -183,7 +188,9 @@ 

开启后,默认交付的是一个读取入口

--select "$POINTER_A" --select "$POINTER_B"

续页引用固定当前选择,沿它返回的 next_command 读至 EOF;更换 Pointer 时,回到原 recovery 引用。极小预算可能保留原显式文件/摘要接口。上面的教学测量也使用这个兼容接口,数字仍对应原实验条件。

-

文件基线和模型此刻保留了什么,需要分别判断。保存文件可核验,不能据此假定模型记住了全部原文。

+

文件保存了什么,模型读到了什么

+

差分编码要回答的是:用已核验的旧文件,能否准确还原当前文件? 原生 captured_file 路径核验完整保存的基线、版本和 scope,不以模型先通读旧文件、某组选读的 EOF 或一次 ACK 作为前提。每份基线本身是完整版本,也不要求模型先追完一串差分。

+

渐进式读取回答另一个问题:这次决定需要哪些事实? 仍以语音示意为例,下一轮只新增了样本 3,隐私约束没有变化。用一个遗漏读取的反例检验:假设 Agent 漏读了必需的隐私约束,或压缩后已经无法确认它。准备外发音频前,仍须从当前恢复入口补读 /goal/constraints,并遵守原有授权边界。它可以直接读这处事实,无须先通读旧基线。文件已保存、字段未变化,都不能代替这次必要读取。

压缩或读取失败,恢复上下文,也保留原工作

宿主发生 compaction,后续新工作轮次建立新的完整基线。原目标、要求和来源仍然存在,必要事实可以重新读到。

如果压缩、缺页或输出截断发生在尚未结束的 Turn 内,就沿原捕获补回缺失事实,保留原任务、执行和结算身份。假设语音结果已经发送,只是回执没有读到,恢复上下文不能变成再发一次结果。

@@ -199,7 +206,7 @@

这些数据说明了什么

这组例子把三个收益来源分开了:少重复搬当前用不上的历史,少发本可合并的读取请求,少重复交付已经核验的不变结构。 完整来源、要求与恢复路径贯穿其中。

上面的 reader 数字包含实际 CLI 响应中的分页元数据;本机路径长度会影响少量封装字节。原生投影数字比较捕获正文,周期表比较编码层,三种口径分别列出。示例脚本可复算这些值与完整还原结果。

放回真实长程任务,还要把必要补读、恢复、缓存命中、模型输入输出、工具调用、等待与人工介入一起算进去;效果则看是否漏掉约束、做错决定,最后交付了什么。这个合成例子测清了机械路径,真实模型收益仍需可比任务验证。

-

当前实现已覆盖完整来源、文件入口、批量选择、目录深度、适用验收分层、精确共享、周期差分与恢复。更广的候选、偏好、Explore 按决策阶段统一交付及 C1 全入口收敛仍有未完成部分。受支持的 Codex 会话未保存偏好时默认开启;在「设置 → 本机 → 能力 → 差分上下文 → Codex」关闭,已有 off 会保留。单次命令或新 benchmark trial 可用 --context-delivery off;trial 不继承本机开关,已有冻结 trial 也不被重写。设置、准确 hooks/配置、实际采用分别核验;hooks 不可用时完整交付。普通 CLI、不支持的宿主和非 heartbeat benchmark 保留原完整路径。具体 2.2 发布资格沿最终版本确认。

+

2.2.0 已交付完整来源、文件入口、批量选择、目录深度、适用验收分层、精确共享、周期差分与恢复。更广的候选、偏好、Explore 按决策阶段统一交付及全入口一致性仍有未完成部分。在「设置 → 本机 → 能力 → 差分上下文 → Codex」可以关闭,已有 off 会保留;单次命令或新 benchmark trial 可用 --context-delivery off。Trial 不继承本机开关,已有冻结 trial 不被重写。默认策略下,普通 CLI、不支持的宿主和非 heartbeat benchmark 保留原完整路径;显式请求不支持的 on 会明确拒绝,hooks 不可用时使用完整交付。保存配置、实际安装与模型采用分别核验。

用同一份材料动手读一遍

example.py 完整生成这个合成语音任务,调用现有 reader、渐进投影与差分 owner,逐项核对选择值、目录地址、全部 10 条要求、共享引用与完整恢复。程序不注册 Goal,不做准入或结算,也不调用模型或语音服务。

先准备原测量的源码 a4a0fad889d91cba0d945ce658788bdb76ba351e 及其兼容依赖,在该 checkout 根目录运行。将下载的程序路径填入下方;输出目录必须为空。

@@ -208,6 +215,6 @@

用同一份材料动手读一遍

生成的 capture.json 保留全部 80 条历史;projection.json、required-context.json 对应完整/必要视图;cycle-base.json、cycle-delta.json 对应周期编码;manifest.json 固定文件与摘要。生产中使用当前返回的真实路径,不能照抄教学树或另一份捕获的数组位置。

原测量记录固定本文数据。程序在其他源码上可以检查机制,但新结果不能改写原观察;页面封装字节还受本机路径长度影响。无需启动一次新研究来理解这个例子。

-

来源:完整交付契约 · 渐进投影 · 选择与目录 · 差分与恢复。例子运行源为 a4a0fad;上述三项核心 owner 与 0e78a19 逐字一致。原生最终交付还会核算引用和后续读取,不能由编码层数字替代。

+

来源:完整交付契约 · 渐进投影 · 选择与目录 · 差分与恢复。例子运行源为 a4a0fad;原测量的三项核心 owner 与 0e78a19 逐字一致;上面的实现链接固定到发布版 v2.2.0,不把旧数字改记为发版后的重测。原生最终交付还会核算引用和后续读取,不能由编码层数字替代。

LoopX

让工作持续推进,让判断由人掌握。

↑
diff --git a/apps/presentation/site/public/blog/zh/edgebench-feedback-and-memory/index.html b/apps/presentation/site/public/blog/zh/edgebench-feedback-and-memory/index.html index 88961448c7..095f7e2722 100644 --- a/apps/presentation/site/public/blog/zh/edgebench-feedback-and-memory/index.html +++ b/apps/presentation/site/public/blog/zh/edgebench-feedback-and-memory/index.html @@ -57,7 +57,8 @@

长任务没做完,何时检查方向?

7. 控制面自己花掉了多少成本?

另一个错误假设是:入口 packet 变短,Agent 的总成本就会下降。把目标、事实、写回说明从入口挪到第二条命令,确实缩短了首屏;模型随后还要读取它们。若中途丢失上下文,又要重建同一份来源。只量入口,就把成本移出了表格。

#7269 将原有捕获交给共享的上下文 reader:先读完整目标、事实和反馈,再读写回格式及结算步骤。review 与 writeback 是同一个不可变决策的两个视图,不是重新做一次准入,也不各自拥有一套权限判断。源码回归逐项核对目标、规划范围、证据引用、完整写回合同和 action signature;小入口必须仍能到达完整义务。

-

差分还增加了一个前提:接收者确实保留了那份精确基础。当前 native Hook 路径在同一会话和 context epoch 内,把完整分页后的正常工具交付作为连续性工程假设。会话变化、已观察到的压缩、缺页或已知交付失败会使它失效,随后读取原来的完整 capture;另开一次 guard 取得的新决策不能代替恢复旧决策。漏掉的 Hook 和不可见的下游截断仍是限制,读取回执不是模型保留内容的独立证明。

+

早期差分路径还依赖一个前提:接收者保留了那份精确基础。下述历史记录中的 native Hook 路径,在同一会话和 context epoch 内,把完整分页后的正常工具交付作为连续性工程假设。会话变化、已观察到的压缩、缺页或已知交付失败会使它失效,随后恢复原来的完整 capture;另开一次 guard 取得的新决策不能代替恢复旧决策。漏掉的 Hook 和不可见的下游截断仍是限制,读取回执不是模型保留内容的独立证明。

+

已发布的 2.2.0 原生文件路径另行核验完整保存的基线:文件能否准确还原,与模型此前是否读过它分开判断。即使返回 delta,Agent 也可通过当前恢复引用直接选读决定需要的事实,不必先通读旧基线。这个区别改变了机械交付的前提;它不会为下面的历史实验补做资格。语音示例把两种连续性放在同一个具体决定里解释。

同一 PR 的一组历史离线对照把所有必读阶段重新加总。以下固定比较基线 59464f974 与测试 head a006a2bea1、同四份保留观察的交付;它不是最终合并 head 的重新测量。

Complete read pathUTF-8 bytes: base → headChange
CLI entry and all exposed detail352,039 → 361,430+2.67%
@@ -110,6 +111,6 @@

把结构输入单独计量

截至 10 月 28 日的上下文研究让 Grid、Schema、Ad Placement 从原始任务开始,使用 Sol/max,每题最多 18h。可用的历史 Official 或旧 LoopX 作为描述性参考,保留模型/effort、源码、评分器、机会和时钟差异,不冒称严格配对;旧材料无法回答相应问题时再补 base。各次尝试按实际源码、连续性依据和传输路径分别分析,不合并成同一处理。

先核实实际交付确实改变:首次完整读取、声明连续性依据的后续差分、已知基线失效时的完整回退,以及一次从复盘到工作的完整交互。再比较有效进展、成果保留和完整用量。开关、回执、局部小包都不能替代这些证据。Feature 偏离预期时,停止相关尝试,保留其成本和历史,修复后用新的尝试身份重启。

新原生 Hook 路径的首批 Grid、Ad 交互已完成完整分页、一次差分及后续任务工作。同一次捕获内,差分载荷分别比对应完整包小 19.53%、19.48%;这只是一轮读取的载荷比较,不含分页封装、后续模型输入、Hook、恢复或复盘费用,不能外推为总成本下降。完整复盘到后续原生成果、持续收益和故障恢复仍待验证。

效果目标是很多小时之后仍能取得有效进展并保住成果;效率目标是用更低完整成本和人工介入取得同等验收价值。下一份结果需要同时回答它们;这些尝试沿用各自原定截止时间,替换尝试的失败和消耗也保留。

-

这里保留的是 10 月 28 日的研究检查点。后续 reader、默认策略或编码器合入,不会使这些未验证项自动通过。机制文章的合成字节测量与本节的模型轨迹属于不同证据层,不能拼成一项节费结论。

数值分别保留各历史模型、版本、窗口与有效性边界。公开实现:LoopX。早期详细记录保留在上一版文章。这里不发布原始轨迹、隐藏评测材料、运行时凭证或私有操作上下文。

+

这里保留的是 10 月 28 日的研究检查点。2.2.0 的正式发布和默认开启策略,不能让这些未验证项自动通过。机制文章的合成字节测量、发布路径的安装/恢复检查、本节的真实模型轨迹,分别回答机械交付、产品可用性与研究效果问题;版本、窗口和指标不同,不能拼成一项节费结论。

数值分别保留各历史模型、版本、窗口与有效性边界。公开实现:LoopX。早期详细记录保留在上一版文章。这里不发布原始轨迹、隐藏评测材料、运行时凭证或私有操作上下文。

diff --git a/apps/presentation/site/public/blog/zh/typescript-control-plane-refactor/index.html b/apps/presentation/site/public/blog/zh/typescript-control-plane-refactor/index.html index ee9c54f32d..1bd52d0d0c 100644 --- a/apps/presentation/site/public/blog/zh/typescript-control-plane-refactor/index.html +++ b/apps/presentation/site/public/blog/zh/typescript-control-plane-refactor/index.html @@ -250,6 +250,11 @@

10. CLI 输出约束:压缩前先保护决策语义

自然语言还需要逐条看:义务、顺序、证据来源、权限限定、适用范围、继续和停止条件。删除一句“仅在原证明成立时”可能比改一个字段更影响行为。

一个有界实例是去掉 JSON 缩进与分隔空白。在同一历史 fixture 的 36 个 Todo、12 条 run 上,输出从 48,418 降到 39,726 字符,完整 payload 解码保持一致,字符数减少约 18%。它证明展示体积变化;token、费用和 Agent 效果需要另测。

预算保护的是消费者能用的接口。新增指导确实有用,可以带同条件证据提高预算;为了变绿而删掉决定所需的条件,反而损坏接口。

+

可恢复的文件,也不能当成模型的记忆

+

前面的恢复问题还在上下文交付中出现:一张旧回执不能续租,一份保存在磁盘上的完整基线也不能证明模型已经读懂其中的约束。

+

2.2.0 的原生文件差分把这两层分开:编码器核验同一有效周期内的完整文件、版本与 scope,准确还原当前结构;Agent 则从当前完整视图,渐进式读取本次判断需要的目标、要求与证据。即使收到 delta,也能直接选读当前事实,不必先读完整旧基线。压缩发生在未结束的工作中时,补回缺失事实,同时保留原任务与结算身份。

+

这项边界让“文件可还原”和“模型掌握了什么”各自可检查。具体的语音任务、JSON Pointer 与默认文件路径,见渐进与差分交付的完整例子。文件与读取路径成立,仍需继续检验它们是否改善了真实研究。

+

11. 最后退役:保留贡献,核对删除的究竟是什么

9 月下旬已经删除了未使用的 authority prototype,以及实验性 Todo event overlay、dual-writer 和 backfill;其中有数据的旧来源有显式保留或拒绝处理。这里没有删除所有 event logging,也没有取消新 authority 的原子事件与回执。

10 月继续隔离普通 Markdown writer,验证模块不在时真实后续写入仍走正确 owner。等 consumer 路由与替代路径接上,才具备集中删除的条件。

完整读取路径UTF-8 字节:基线 → head变化
CLI:入口与全部暴露详情352,039 → 361,430+2.67%