billion-context-dsh 0.2.1 → 0.2.3

package/README.en.md

@@ -3,7 +3,7 @@
 [English](./README.en.md) | [中文](./README.md)
 
 > **⚠️ Beta notice — not for production use**
-> This project (**v0.2.1**) is a work-in-progress beta. The [DeepSeek Harness](https://github.com/deepseek-ai/deepseek-harness) itself is also in **public beta**. **Do not use either in engineering / production environments** — expect breaking changes and rough edges.
+> This project (**v0.2.3**) is a work-in-progress beta. The [DeepSeek Harness](https://github.com/deepseek-ai/deepseek-harness) itself is also in **public beta**. **Do not use either in engineering / production environments** — expect breaking changes and rough edges.
 
 <p align="center">
 <strong>Built with gratitude on top of these projects</strong> — please give them a ⭐:
@@ -41,10 +41,11 @@ When conversations get long, the model runs out of context. Most tools hard-trun
 
 Unlike DSH's built-in auto-compaction (which replaces a range with an automatically generated summary), billion-context-dsh:
 
-- **Model-driven** — the model writes the summary itself; there is no second LLM summarization call (the ACP cost win)
+- **Model-driven** — the model writes the summary itself; there is no second LLM summarization call
 - **Advisory, never imperative** — automatic policy only *nudges*; the model decides whether and when to compress
 - **Durable & recoverable** — a compressed range becomes a checkpoint node, the originals stay in the append-only session log; `decompress` restores them, `search_context` finds information inside blocks
-- **Seq-based refs** — no message tags; surface seqs are carried by the nudge's range table, with auto-balanced range edges and `#callId` tolerance
+- **Long tasks hold steady** — every step builds on the results before it; key conclusions stay usable and compound, so very long tasks actually finish
+- **Context stays lean** — every request rides on a small, distilled slice of context with only the key information; no bulk compression of large ranges, so details don't decay with it — and tokens stay low
 
 This is the DeepSeek Harness port of [billion-context-pi](https://github.com/ranxianglei/billion-context-pi) (the Pi coding-agent adapter): the compression core ([acp-kernel](https://github.com/ranxianglei/acp-kernel)) is reused verbatim, and the adapter layer was rewritten against DSH's durable-surface model — see [docs](https://github.com/Tyan66666/billion-context-dsh/tree/main/docs) for the verified mapping.
 
@@ -129,7 +130,7 @@ DSH derives every model request from its append-only session log (the *surface*)
 | nudge ("efficiency note — compress early and keep context lean") | injected at `agent/pre-step` by the kernel's pressure decision — efficiency note + context breakdown + compression rules, tone aligned with kernel/pi; never an order |
 | `decompress` | read-only recovery of shadowed originals from the log |
 | `search_context` | scores block summaries + originals rebuilt from the log |
-| `acp_status` | block ledger + context pressure |
+| `acp_status` | CONTEXT BREAKDOWN (tool/text/summaries shares of the visible total) + compressed-block ledger + nudge decision line; no context-window rows; scope/view/tool/sort/limit drilldown supported |
 | block state | in-memory kernel state + **log-rebuilt ledger** (no sidecar files) |
 | tiered distillation (T2/T3) | re-compressing a block's summary node distills that block (tier 2); distilling a tier-2 block yields tier 3. Tier + kernel block ids are persisted to the log, so kernel state rehydrates from the log after a restart and stays distillable |
 
@@ -146,9 +147,9 @@ A walkthrough of the ACP philosophy this project inherits — how active context
 | Tool | What it does |
 | --- | --- |
 | `compress` | Replace a seq range with a dense summary you write (edges auto-balanced to tool-pair boundaries); re-compressing a block's summary node distills it (tier 2/3) |
-| `decompress` | Restore a previously compressed block's original content (read-only) |
+| `decompress` | Restore a previously compressed block's original content (read-only); accepts the `bN` ref shown by acp_status or a compaction id |
 | `search_context` | Search compressed block summaries and originals by keyword |
-| `acp_status` | Context usage, compressed blocks, compressible ranges |
+| `acp_status` | CONTEXT BREAKDOWN (tool/text/summaries shares of the visible total) + compressed-block ledger + nudge decision line; no context-window rows. Drilldown supported: `scope:"compressed"` per block, `scope:"uncompressed"` + `view:"messages"`/`"ranges"` per message/range, with `tool` filter, `sort` order and `limit` cap. Drilldown row refs are kernel ids (mN, size awareness only) — compress always uses the `Surface:` seqs |
 | `/acp` | status / compress / decompress from the command bar |
 
 ## Upstream & credits
@@ -169,7 +170,7 @@ This project reuses `acp-kernel`'s compression core and `billion-context-pi`'s d
 | Key | Default | Meaning |
 |---|---|---|
 | `modelContextLimit` | auto-detected (fallback `128000`) | Context window used for the kernel's pressure decisions; an explicit value wins and skips the probe |
-| `autoModelContextLimit` | `true` | Probe the model's real window from the model API (`agent.ctx.llm.resolveModelInfo`); fall back to the default on failure, `acp_status` shows the window source |
+| `autoModelContextLimit` | `true` | Probe the model's real window from the model API (`agent.ctx.llm.resolveModelInfo`); fall back to the default on failure, the `/acp` command shows the window source (the `acp_status` model tool carries no window info) |
 | `nudgeMinContextLimitPct` | kernel default `0.45` | Nudge window lower bound (usage fraction) — validation only; the growth-driven trigger has no percentage floor — same default as billion-context-pi |
 | `nudgeMaxContextLimitPct` | engine default `0.70` (kernel/pi default `0.75`) | Over-limit line: above this the nudge fires regardless of growth — deliberately below the host compaction-basic 80% auto-compaction line so the forced nudge fires first; an explicit value wins |
 | `nudgeEmergencyThresholdPct` | engine default `0.85` (kernel/pi default `0.95`) | Emergency nudge (bypasses the per-turn dedup) — lowered from `0.95`: at 95% the model has no room to act and the 80% auto-compaction line shadows it; an explicit value wins |

package/README.md

@@ -3,7 +3,7 @@
 [中文](./README.md) | [English](./README.en.md)
 
 > **⚠️ 测试版声明——请勿用于生产环境**
-> 本项目（**v0.2.1**）仍处于开发中的测试版。[DeepSeek Harness](https://github.com/deepseek-ai/deepseek-harness) 本身也处于**公开测试版**阶段。**请勿将两者用于工程化 / 生产环境**——预期会有破坏性变更与粗糙之处。
+> 本项目（**v0.2.3**）仍处于开发中的测试版。[DeepSeek Harness](https://github.com/deepseek-ai/deepseek-harness) 本身也处于**公开测试版**阶段。**请勿将两者用于工程化 / 生产环境**——预期会有破坏性变更与粗糙之处。
 
 <p align="center">
 <strong>衷心感谢以下项目——请给它们一个 ⭐：</strong>
@@ -41,10 +41,11 @@
 
 与 DSH 内置的自动压缩（用自动生成的摘要替换一段范围）不同，billion-context-dsh：
 
-- **模型驱动** —— 摘要由模型自己书写，没有第二次 LLM 摘要调用（ACP 的成本优势）
+- **模型驱动** —— 摘要由模型自己书写，没有第二次 LLM 摘要调用
 - **只建议、不强令** —— 自动策略只 *nudge*（提醒），是否压缩、何时压缩由模型决定
 - **持久且可恢复** —— 压缩范围成为 checkpoint 节点，原文保留在 append-only 会话日志中；`decompress` 可恢复，`search_context` 可在块内查找
-- **基于 seq 引用** —— 不需要消息标签；surface seq 由 nudge 的范围表携带，范围边界自动平衡、容忍 `#callId` 片段
+- **长任务稳得住** —— 每一步都接着前面的成果走，关键结论持续可用、不断叠加，超长任务更容易跑完
+- **上下文始终精简** —— 每次请求都只用少量、精炼的上下文，只保留关键信息；不做大段统一压缩，细节不随之衰失，token 消耗自然更低
 
 这是 [billion-context-pi](https://github.com/ranxianglei/billion-context-pi)（Pi 编码代理适配器）在 DeepSeek Harness 上的移植：压缩内核（[acp-kernel](https://github.com/ranxianglei/acp-kernel)）原样复用，适配层针对 DSH 的 durable-surface 模型重写——经过验证的映射关系见 [docs](https://github.com/Tyan66666/billion-context-dsh/tree/main/docs)。
 
@@ -130,7 +131,7 @@ DSH 的每个模型请求都派生自其 append-only 会话日志（*surface*）
 | nudge（"效率提示——尽早压缩保持精简"） | 由内核的压力决策在 `agent/pre-step` 注入——效率通知 + 上下文分解 + 压缩规则，语气对齐 kernel/pi；绝非命令 |
 | `decompress` | 从日志只读恢复被遮蔽的原文 |
 | `search_context` | 对从日志重建的块摘要与原文打分 |
-| `acp_status` | 块账本与上下文压力 |
+| `acp_status` | CONTEXT BREAKDOWN（tool/text/summaries 占可见总量）+ 压缩块账本 + nudge 决策行；不含上下文窗口；支持 scope/view/tool/sort/limit 钻取 |
 | 块状态 | 内存内核状态 + **日志重建账本**（无旁车文件） |
 | 分层蒸馏（T2/T3） | 再次压缩某块的摘要节点 = 蒸馏该块（tier 2），蒸馏 tier-2 块得 tier 3；tier 与内核块 id 持久化进日志，重启后内核状态从日志再水合、可继续蒸馏 |
 
@@ -147,9 +148,9 @@ DSH 的每个模型请求都派生自其 append-only 会话日志（*surface*）
 | 工具 | 作用 |
 | --- | --- |
 | `compress` | 用你书写的紧凑摘要替换 seq 范围（边界自动平衡到 tool-call/result 配对点）；对某块的摘要节点再次压缩 = 分层蒸馏（tier 2/3） |
-| `decompress` | 恢复已压缩块的原始内容（只读） |
+| `decompress` | 恢复已压缩块的原始内容（只读）；接受 acp_status 显示的 `bN` 或 compaction id |
 | `search_context` | 按关键词搜索压缩块摘要与原文 |
-| `acp_status` | 上下文占用、压缩块、可压缩范围 |
+| `acp_status` | CONTEXT BREAKDOWN（tool/text/summaries 占可见总量）+ 压缩块账本 + nudge 决策行；不含上下文窗口。支持钻取：`scope:"compressed"` 逐块、`scope:"uncompressed"` + `view:"messages"`/`"ranges"` 逐消息/区间，`tool` 过滤、`sort` 排序、`limit` 截断。钻取行 ref 是内核 mN（仅供体量感知）——压缩始终用 `Surface:` 的 seq |
 | `/acp` | 从命令栏执行 status / compress / decompress |
 
 ## 上游项目与致谢
@@ -170,7 +171,7 @@ DSH 的每个模型请求都派生自其 append-only 会话日志（*surface*）
 | 键 | 默认值 | 含义 |
 |---|---|---|
 | `modelContextLimit` | 自动探测（回退 `128000`） | 用于内核压力决策的上下文窗口；显式配置时优先且跳过探测 |
-| `autoModelContextLimit` | `true` | 从模型 API 自动探测真实窗口（`agent.ctx.llm.resolveModelInfo`）；探测失败回退默认值，`acp_status` 展示窗口来源 |
+| `autoModelContextLimit` | `true` | 从模型 API 自动探测真实窗口（`agent.ctx.llm.resolveModelInfo`）；探测失败回退默认值，`/acp` 命令展示窗口来源（模型工具 `acp_status` 不含窗口信息） |
 | `nudgeMinContextLimitPct` | 内核默认 `0.45` | Nudge 窗口下界（用量占比）——仅作配置校验，增长路径的触发没有百分比下限——与 billion-context-pi 相同的默认值 |
 | `nudgeMaxContextLimitPct` | engine 默认 `0.70`（内核/pi 默认 `0.75`） | 过限线：超过此值则无论增长与否都触发 nudge——刻意低于宿主 compaction-basic 的 80% 自动压缩线，保证强制 nudge 先触发；显式配置优先 |
 | `nudgeEmergencyThresholdPct` | engine 默认 `0.85`（内核/pi 默认 `0.95`） | 紧急 nudge（绕过每轮去重）——从 `0.95` 下调：95% 时模型已无操作空间且会被 80% 自动压缩线遮蔽；显式配置优先 |

package/dist/index.d.ts

@@ -40,7 +40,7 @@ export { makeTools, type ToolEnvironment } from './tools.ts';
 export { acpCommand } from './commands.ts';
 export { buildNudge, resolveTokenCount, type NudgeEnvironment, type NudgeOutcome } from './nudge.ts';
 export { DEFAULT_CONTEXT_WINDOW, detectContextWindow, windowSourceLabel, type AcpWindow, } from './window.ts';
-export { AlreadyCompressedRangeError, rebuildBlockLedger, resolveSurfaceRange, runCompactionTransaction, shadowedSeqsOf, findOpenTurn, assertNoActiveCompaction, blockRegistry, blockRefForSummarySeq, compactionIdsOfKernelBlocks, summarySeqOfKernelBlock, expandShadowedSeqs, type AcpBlockLedgerEntry, type CompactionTransactionInput, type ResolvedSurfaceRange, } from './region.ts';
+export { AlreadyCompressedRangeError, rebuildBlockLedger, resolveSurfaceRange, runCompactionTransaction, shadowedSeqsOf, findOpenTurn, assertNoActiveCompaction, blockRegistry, blockRefForSummarySeq, compactionIdsOfKernelBlocks, summarySeqOfKernelBlock, expandShadowedSeqs, hideCompressToolPair, stripOrphanedSurfaceToolMessages, type AcpBlockLedgerEntry, type CompactionTransactionInput, type ResolvedSurfaceRange, } from './region.ts';
 export { eventsToCoreMessages, projectEvent, surfaceEventsOf, extractEventText } from './messages.ts';
 export interface AcpConfig {
     /**
@@ -108,6 +108,8 @@ export declare class AcpCompactionEngine extends CompactionEngine {
     /** Resolved prompt templates (validated at construction — fail-fast on template typos). */
     readonly prompts: ResolvedPrompts;
     private readonly lastNudgeTurn;
+    /** Successful compress call ids awaiting their tool/result so the pair can be hidden. */
+    private readonly compressCallIdsToHide;
     /** Per provider/model route the resolved window (probe failures cached too). */
     private readonly windowCache;
     constructor(ctx: Context, config?: Partial<AcpConfig>);