[ranxianglei]

Closes #3975

kernel-change: the consumed-call cleanup (kernel/src/hide-consumed.ts) cannot parse the plain-text line-form content that the compress tool itself accepts (parseLineEntry); the host side cannot fix this without duplicating the line-form parser in a second location (format-drift risk — the #3780 precedent already locates the shared home for this decode in the kernel). Issue #3975 is the dedicated issue naming the kernel root cause.

Rebased onto current master (incl. #3769, which replaced the old keep-last-two orphan rotation with the fold-gated sweep); tests adapted to the current semantics, end-state guarantees unchanged.

What breaks today

parseCallText understood object-array content and stringified JSON arrays only. A plain-text line-form content string (first line m00100–m00120 topic, remaining lines the summary) resolved to no quoted ranges, so such calls mis-resolved as orphans: on quiet turns they stayed visible unstubbbed — retaining their full summary text forever alongside the kernel acp_summary — and on fold turns they were swept as failure residue although they ARE the folded block's summary carrier (#1567). Measured in the reporter's snapshot: 2 of 15 mcp__bili__compress calls unsimplified (~2.68K tokens args, ~2.54K redundant full-text overlap across 10 active blocks).

The same mechanism also kept bare-name proxy-lane line-form calls at full length forever (id-matched fate always keeps; the rewrite returned null pre-fix).

Fix

Reuse the tool's own acceptance ladder instead of a second reader:

• parse-compress-input.ts: new exported parseLineContent(content) — stripJsonWrapperResidue → splitLineEntries → parseLineEntry, i.e. exactly what the compress tool accepts when content is not a JSON array. Unparsable entries keep source/header only and never claim a range.
• hide-consumed.ts: string content whose inner JSON value is not an array (or fails to parse) flows through parseLineContent; quoted-range keys use the normalized refs (aligning with how line-form specs store block refs); rewrite/compact rebuild the line form with headers verbatim and long summaries stubbed to 200 chars, re-emitting the full args object with the string shape preserved (fix: persistent overlay homes for pi/omp/hermes launchers (session history loss) #230 convention). Garbage/refless strings stay byte-identical.

Line-form calls now resolve active/inactive by range exactly like object-array calls and get stubbed identically; on fold turns they are fate=active and therefore exempt from the #3769 sweep, while true orphans are still swept at/after the fold start.

Behavior change (old → new)

Surface	Old	New
Rendered args of a line-form compress call with long summaries	full text retained forever (both MCP lane and bare-name proxy lane)	long summaries stubbed to 200 chars, header lines verbatim — identical end state to object-array calls
Fate of line-form calls	always orphan (full text visible on quiet turns; swept as failure residue on folds despite being the block's carrier)	active/inactive by quoted range; quiet-turn orphans still visible but stubbed
Kernel acp_summary retention for these blocks	unchanged (anchor stays the sole complete carrier)	unchanged — deliberately NOT extended: post-stub args are not a complete carrier, so stripping the anchor would lose the full text (issue expectation #3)

Object-array, stringified-array, and garbage paths are regression-pinned unchanged.

Config surface

None — no config/env/CLI changes.

Tests

• kernel/tests/issue3975-hide-consumed-lineform.test.ts (new, 9): normalization (m100→m00100, b2/b5, topic), refless-garbage verbatim, MCP synthetic-id plain-text carrier stubbed as STRING, fold-sweep exemption for a live-range plain-text carrier while a true orphan is swept ([Bug] MCP failed compression record rotation prematurely rewrites the history prefix, triggering expensive cache recompu | [Bug] MCP 失败压缩记录轮换提前改写历史前缀，触发大额缓存重算 #3769/plugin mode: in-place acp_summary never stripped when the client's own compress pair rides inbound — double summary carrier on every post-fold turn #1567), partial-range invalidation (live kept+stubbed / dead dropped), all-dead hidden, quiet-turn failed-fold orphans visible but stubbed, bare-name id-matched lane, stringified-array shape+stub regression (fix: persistent overlay homes for pi/omp/hermes launchers (session history loss) #230).
• tests/issue3975-mcp-lineform-endstate.test.ts (new, host level): issue expectations 2–3 pinned end-to-end — after hideConsumedCompressCalls + stripKernelSummaries, the plain-text call ends stubbed while its acp_summary_b0 anchor stays byte-complete; object-array parity pin.
• Kernel suite 1407/1407 green locally; targeted bili lanes green (25/25); full bili suite = CI.

[github-actions]

📦 Built Package Artifact

Branch: wt-3975-ework-daemon (cedd2a4)

Option A — Install this exact build (recommended)

npm install -g billion-context@0.0.4817-pr.4032.1512ad8

This exact version is immutable and reproducible. The rolling pr tag tracks the newest test build across all PRs.

Option B — Download artifact

1. Download the artifact from the Actions run
2. Extract the tarball and install:

tar xzf billion-context-pr4032.tgz
npm install -g package

---

Install instructions are refreshed on each push.

[ranxianglei]

🤖 Powered by ework · qwen3.8-27b

(ai: vllm-qwen/qwen3.8-27b)

查重：不重复。 本 PR 是 #3975 的专属修复 PR（Closes #3975）。相邻条目核对：#3785（line-form 双重转义换行打捞，已合）改的是解析入口的容错，不是 consumed-call 清理；#2970/#2977（MCP compress 路由进清理层）、#3769/#3784（orphan 清扫门控）是本 PR 依赖的既有机制而非平行修复；#4311（摘要双份保留，已关）与本 PR 同症状家族但指向的是载体去重边界，PR 已按该边界刻意不扩大 carrier 索引。无并列的同机制在途 PR。

标签：pr + bug + scope/kernel + p2（缺陷类；根因与改动都在 kernel 的 consumed-call 清理面；能用但有冗余驻留、有绕行——p2；MCP shim 为多宿主共用面，不打具体客户端标签）。

开始独立审核（不照抄 #3975 线程结论）：拉分支通读 diff、清点触碰路径的既有行为、本地定向跑新增测试与相关套件、核对 CI。另注意到 PR head 相对当前 master 显示 merge conflict（master 已推进约 3 天），审核中一并处理 rebase。结论随后按八问回复。

[ranxianglei]

🤖 Powered by ework · qwen3.8-27b

(ai: vllm-qwen/qwen3.8-27b)

PR #4032 独立复审结论（对最新 master 重读 diff 与行为，不复述 #3975 线程结论）。

已执行的动作

• rebase 到最新 master（期间 master 前进了 ~224 commits，含 [Bug] MCP failed compression record rotation prematurely rewrites the history prefix, triggering expensive cache recompu | [Bug] MCP 失败压缩记录轮换提前改写历史前缀，触发大额缓存重算 #3769/compress calls rejected when the summary text rides in content next to top-level startId/endId — plus U+2212 dash silently folds the wrong range #4252/[Bug] compress range straddling the protected latest user message: summary of the reply is anchored BEFORE that user message (causal order inversion) #3953/Accident re-anchor debris is unrecoverable: the overlay island lands mid-turn, and the turn-split guard means the orphan | 事故重锚碎屑无法回收:覆盖岛落在 turn 中段,turn-split 守卫让孤另一半永远折不了 #4348）；冲突仅版本行，按 Kernel Boundary 规则解为 kernel 0.0.159→0.0.160，两个 lockfile 仅版本行变化（已核对 delta）；
• 补 2 个回归钉（commit 40c9d06a0）：① clamped-range compact fallback——id-matched 调用引用的范围与块实存范围不一致时（[Bug] compress range straddling the protected latest user message: summary of the reply is anchored BEFORE that user message (causal order inversion) #3953 内部受保护消息截断），live-range rewrite 保留为空、compact 兜底仍须 stub，此前无测试覆盖该可达路径；② stub 幂等定点——第二遍处理已 stub 的文本输出字节一致（跨轮缓存稳定的直接证据）；
• 本地验证：typecheck 干净（含 kernel/pi-subagents 构建）；定向测试 163/163 绿（kernel：issue3975-lineform(12)/hide-consumed-mcp/hide-consumed-stub/issue3769-orphan-sweep/parse-compress-input/orphan-fixes/orphan-endpoint-snap/prune-orphan-mcp/hide-merge-rebuild/issue3371；host：issue3975-mcp-lineform-endstate）；全量套件以 CI 为准；
• 已推送 wt-3975-ework-daemon（40c9d06a0），PR 冲突解除（mergeable=true）。CI 12 条流水线已触发，当前在 runner 队列等待（并行会话多，排队中）。

八问

1. 是什么问题：compress 工具接受纯文本 line-form content（首行 m00100–m00120 主题），但 consumed-call 清理只认对象数组/字符串化数组——line-form 调用被误判孤儿：完整摘要永不被截成 200 字符，与 kernel acp_summary 长期重复占上下文；且 fold 时被当「失败残留」扫掉，尽管它正是该块摘要的 carrier。
2. 什么原因导致的：清理层（hide-consumed 的 parseCallText）是工具的第二个读取器，没有复用工具自己的接受阶梯（stripJsonWrapperResidue→splitLineEntries→parseLineEntry），两个读取器漂移，line-form 形状缺失。修在根上=单一解码源归 kernel（compress: literal backslash-n in string content still fails as kind=content-not-array, hiding the real missing-summary reason #3780 先例：共享解码的家在内核）。
3. 影响范围与严重程度：所有用 line-form compress 的会话（MCP 名车道 + bare-name 代理车道均命中，后者 id-matched 也永不 stub）。报告者实测 15 次调用中 2 次未简化（~2.68K tokens args、~2.54K 重复重叠）。严重程度 p2：可用但持续冗余上下文成本，无数据丢失。注意 compress calls rejected when the summary text rides in content next to top-level startId/endId — plus U+2212 dash silently folds the wrong range #4252 之后 line-form 是 schema 唯一advertised 形状，新流量主力即此形，修复重要性相应提高。
4. 怎么修复的：kernel 导出 parseLineContent，逐字走工具自己的接受阶梯；hide-consumed 把「内层 JSON 非数组或解析失败」的字符串内容路由给它；range key 用规范化 refs；rewrite/compact 按 line-form 重建（header 逐字、>200 字符摘要截 200、字符串形状保留，fix: persistent overlay homes for pi/omp/hermes launchers (session history loss) #230 约定）；垃圾/无 ref 字符串保持字节不变。fate/sweep 逻辑本体未动，只扩展消费点。
5. 是否有回归问题：逐项清点既有行为——对象数组/字符串化数组/垃圾路径回归钉住不变（fix: persistent overlay homes for pi/omp/hermes launchers (session history loss) #230 字符串形状约定、无 ref 垃圾字节稳定均有测试）；[Bug] MCP failed compression record rotation prematurely rewrites the history prefix, triggering expensive cache recompu | [Bug] MCP 失败压缩记录轮换提前改写历史前缀，触发大额缓存重算 #3769 fold-gated sweep 语义保持（测试同时钉 carrier 豁免 + 真孤儿清扫两方向）；quotesActiveBlock 不解析 line-form，判定与对象形 clamped-range 行为对称（都落到 fold 时处理），非回归；acp_summary 锚点保留刻意不扩展——stub 不是完整载体，删锚会丢全文（issue 期望 Use kernel renderTags:"text-only" + forbid tag echoing in prompt #3）；无 fail-fast 变吞错，SUMMARY_STUB_CHARS=200 常量未动，无阈值/时序/输出格式漂移。两种压缩模式都过了一遍：清理层作用于重发历史，MCP/bare-name 双车道各有测试钉。
6. 是否破坏缓存：否。重建确定性（同输入同字节，JSON.stringify 键序稳定），stub 输出是定点（新增幂等测试实证第二遍字节不变），前缀缓存稳定。
7. 是否改了设计方向：否。方向=单一共享解码源放 kernel，与 compress: literal backslash-n in string content still fails as kind=content-not-array, hiding the real missing-summary reason #3780/[Bug] recommend/apply 对同一区间的 turn-integrity 判定永久不一致：acp_status 反复推荐的区间被 apply 侧整段拒绝（0.1.196，Range would split N turn(s) at the protected-zone boundary） #3236「判定与执行共用单一投影」同源。
8. 是否引入新配置：无。

已知残留（不阻塞，已在 #3975 记为 checklist 项）：parser 仍接受的两种 legacy flat 形状（{startId,endId,summary} 无 content；{content: 摘要文本, startId, endId}，#4252 后不再 advertised）cleanup 同样不解析——前者在其块 active 期间全程全文可见（从不 stub），后者下次 fold 正常被清（只是中间不 stub）。两者均先于本 PR 存在、属未 advertised 的衰减形状、无数据丢失；是否扩展 cleanup 覆盖留给 owner 决策。

判定：代码闭环、测试有牙（10+2 kernel / 2 host，含负向门）、冲突已解、本地定向证据齐。等 CI 全局门转绿后建议合并——kernel 变更属 human-only review 档，请人工合并：#4032 。CI 结果出来我再补一条确认。