Coder/deepseek-harness
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

docs: reserve seam for complete capabilities

Browse Source
This commit is contained in:
Turtle
2026-08-09 15:34:32 +08:00
parent 27ac49e687
commit dda02250f5
966 changed files with 2166 additions and 2159 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
.agents/notes/proposed/feature/2026-06-30-pre-tool-input-rewrite.i18n.yaml
View File

@@ -2,5 +2,5 @@
# side as of the last confirmed-consistent state. Both languages carry equal authority;
# after editing either side, bring the other along and re-record with:
# pnpm run verify-translation-pairing --write .agents/notes/proposed/feature/2026-06-30-pre-tool-input-rewrite.md
2026-06-30-pre-tool-input-rewrite.md: f35e6af465ce8cec5685911c43e12b8dd66f2e6a
2026-06-30-pre-tool-input-rewrite.zh.md: 5793b38293e117e328350722ad1efb3126fd176c
2026-06-30-pre-tool-input-rewrite.md: 63027079a996f2f7b9d2ef87afcd4e64a2691fb1
2026-06-30-pre-tool-input-rewrite.zh.md: 8930a3a0b1ca136eb2c788e081f4093df8940162

4
.agents/notes/proposed/feature/2026-06-30-pre-tool-input-rewrite.md
View File

@@ -6,7 +6,7 @@ English | [中文](2026-06-30-pre-tool-input-rewrite.zh.md)
## Problem
The [interception-seams Agent Note](../../implemented/feature/2026-06-30-interception-seams.md) defines `tools/pre-execute` as an allow/deny/ask gate over an execution whose identity is already protected and whose arguments are deeply frozen. Claude Code's `PreToolUse` hook also offers `updatedInput`, so a faithful bridge needs an explicit rewrite mechanism. A rewrite cannot be a mutation escape hatch on the existing execution object: it must keep the durable history, audit record, presentation, and executed value consistent.
The [interception extension-points Agent Note](../../implemented/feature/2026-06-30-interception-extension-points.md) defines `tools/pre-execute` as an allow/deny/ask gate over an execution whose identity is already protected and whose arguments are deeply frozen. Claude Code's `PreToolUse` hook also offers `updatedInput`, so a faithful bridge needs an explicit rewrite mechanism. A rewrite cannot be a mutation escape hatch on the existing execution object: it must keep the durable history, audit record, presentation, and executed value consistent.
## The problem: three readers of pre-execution arguments
@@ -49,5 +49,5 @@ Allowing a pre-execute listener to assign `exec.arguments` would provide only an
- Does rewriting the `assistant/message` tool-call block corrupt any provider's expectation on replay, or is a separate correction safer?
- Should the original arguments be preserved on the `tool/call` event (audit) and, if so, under what field?
- Does the rewrite decision move before the log commit or become a dedicated earlier seam, and how do existing pre-tool allow/deny hooks avoid running twice?
- Does the rewrite decision move before the log commit or become a dedicated earlier extension point, and how do existing pre-tool allow/deny hooks avoid running twice?
- How does this interact with a future permission `ask` flow (a user approving a rewritten call)?

4
.agents/notes/proposed/feature/2026-06-30-pre-tool-input-rewrite.zh.md
View File

@@ -6,7 +6,7 @@ Status: proposed
## 问题
[拦截 seam Agent Note](../../implemented/feature/2026-06-30-interception-seams.md) 将 `tools/pre-execute` 定义为一道针对执行的允许/拒绝/询问门禁，此时执行的身份标识已受保护、参数已被深度冻结。Claude Code 的 `PreToolUse` 钩子还提供了 `updatedInput`，因此忠实的桥接需要一个显式的重写机制。重写不能是对现有执行对象的可变逃逸口：它必须保持持久化历史、审计记录、展示层与实际执行值之间的一致性。
[拦截扩展点 Agent Note](../../implemented/feature/2026-06-30-interception-extension-points.md) 将 `tools/pre-execute` 定义为一道针对执行的允许/拒绝/询问门禁，此时执行的身份标识已受保护、参数已被深度冻结。Claude Code 的 `PreToolUse` 钩子还提供了 `updatedInput`，因此忠实的桥接需要一个显式的重写机制。重写不能是对现有执行对象的可变逃逸口：它必须保持持久化历史、审计记录、展示层与实际执行值之间的一致性。
## 问题本质：执行前参数的三个读取方
@@ -49,5 +49,5 @@ Status: proposed
- 重写 `assistant/message` 中的工具调用块是否会破坏某些提供方在回放时的预期？还是单独的修正更安全？
- 原始参数是否应保留在 `tool/call` 事件（审计）上？如果是，放在什么字段？
- 重写决策是移到日志提交之前，还是成为一个专门的更早 seam？现有的 pre-tool 允许/拒绝钩子如何避免运行两次？
- 重写决策是移到日志提交之前，还是成为一个专门的更早扩展点？现有的 pre-tool 允许/拒绝钩子如何避免运行两次？
- 这与未来的权限 `ask` 流程（用户批准一个被重写的调用）如何交互？

4
.agents/notes/proposed/feature/2026-07-06-recallable-compaction.i18n.yaml
View File

@@ -2,5 +2,5 @@
# side as of the last confirmed-consistent state. Both languages carry equal authority;
# after editing either side, bring the other along and re-record with:
# pnpm run verify-translation-pairing --write .agents/notes/proposed/feature/2026-07-06-recallable-compaction.md
2026-07-06-recallable-compaction.md: 6f7baedeaae0038b7f5106cd5fc4e23bdfbbd8dd
2026-07-06-recallable-compaction.zh.md: cbf8b4a3843ca35647cc569255dfe9d26b45b988
2026-07-06-recallable-compaction.md: 5fd26dd2d4086dd6f834f7bdb3194b0ad16db1eb
2026-07-06-recallable-compaction.zh.md: 25ce6ca50d1755e3a90dbadb35d027c95262e5a4

2
.agents/notes/proposed/feature/2026-07-06-recallable-compaction.md
View File

@@ -83,7 +83,7 @@ Specified during review, deferred until observation calls for them:
- **Pure stubs, no state checkpoint** — rejected: presumes the model knows what it is missing; fails on unknown unknowns.
- **LLM aging/consolidation of frozen chunks** — rejected as a routine mechanism: summary-of-summary loss and frozen-prefix churn; the code-only rollup is its surviving form, deferred.
- **Full prefix as chunk-summarizer input** — rejected: O(N²); the state document gives the same background at O(state).
- **One summarize call emitting all outputs** — rejected: the summarize path has no structured-output enforcement; parsing one free-text response apart is the fragile seam the fail-closed design avoids.
- **One summarize call emitting all outputs** — rejected: the summarize path has no structured-output enforcement; parsing one free-text response apart is the fragile boundary the fail-closed design avoids.
- **Model-chosen chunk boundaries** — deferred: parse-and-validate cost against unproven value; chunk policy sits behind config.
- **Model-authored pointers** — rejected: pointers must be exact; deterministic assembly is.
- **FTS/vector index sidecar** — rejected in-session: the live log is in memory and bounded, a literal scan under budget suffices; an index earns its keep at cross-session scope.

2
.agents/notes/proposed/feature/2026-07-06-recallable-compaction.zh.md
View File

@@ -83,7 +83,7 @@ Status: proposed
- **只保留纯存根，不设状态检查点**：不予采纳，因为这假定模型知道自己缺少什么，在面对未知的未知时会失败。
- **由 LLM 老化／整合冻结分片**：不作为常规机制，因为摘要的摘要会丢失信息，并使冻结前缀频繁变化；其保留下来的形式是由代码汇总，且延后实现。
- **把完整前缀作为分片摘要器输入**：不予采纳，因为成本为 O(N²)；状态文档以 O(state) 提供相同背景。
- **一次摘要调用输出全部结果**：不予采纳，因为摘要路径没有结构化输出约束；解析一份自由文本响应并将其拆开，正是保守失败设计要避免的脆弱 seam。
- **一次摘要调用输出全部结果**：不予采纳，因为摘要路径没有结构化输出约束；解析一份自由文本响应并将其拆开，正是保守失败设计要避免的脆弱边界。
- **由模型选择分片边界**：延后实现，因为相对于未经证明的收益，解析与校验成本过高；分片策略由配置控制。
- **由模型编写指针**：不予采纳，因为指针必须精确，应由确定性代码组装。
- **FTS／向量索引伴随存储**：在会话内不予采纳，因为实时日志已在内存中且大小有界，在预算内进行字面量扫描已经足够；只有跨会话范围才能证明索引的价值。

4
.agents/notes/proposed/feature/2026-07-08-interactive-side-sessions.i18n.yaml
View File

@@ -2,5 +2,5 @@
# side as of the last confirmed-consistent state. Both languages carry equal authority;
# after editing either side, bring the other along and re-record with:
# pnpm run verify-translation-pairing --write .agents/notes/proposed/feature/2026-07-08-interactive-side-sessions.md
2026-07-08-interactive-side-sessions.md: 228380ca2d3d7b2efeb37e713574e1863ce17f38
2026-07-08-interactive-side-sessions.zh.md: f4e3bfdddff16fe687f2637b3626e7e1382eba61
2026-07-08-interactive-side-sessions.md: ff7f152b8a106c5b3856846e6980160d1ed59b43
2026-07-08-interactive-side-sessions.zh.md: f81a3809231ac18ba6b134abd4a821f47af8b662

2
.agents/notes/proposed/feature/2026-07-08-interactive-side-sessions.md
View File

@@ -36,6 +36,6 @@ Rewind productization, session-tree views, a model-facing side-session tool, and
## Risks
- Read-only behavior is advisory until a `tools/pre-execute` deny gate enforces it; [the interception seam](../../implemented/feature/2026-06-30-interception-seams.md) can add that gate without changing these mechanics.
- Read-only behavior is advisory until a `tools/pre-execute` deny gate enforces it; [the interception point](../../implemented/feature/2026-06-30-interception-extension-points.md) can add that gate without changing these mechanics.
- A compacted source forks its compacted view, so a bound surface should disclose that the child inherits summaries rather than replaced turns.
- Repeated handbacks consume parent context. The per-merge length cap bounds each note; later consolidation belongs to compaction.

2
.agents/notes/proposed/feature/2026-07-08-interactive-side-sessions.zh.md
View File

@@ -36,6 +36,6 @@ Status: proposed
## 风险
- 只读行为在 `tools/pre-execute` 拒绝门禁强制执行之前仅为建议性质；[拦截 seam](../../implemented/feature/2026-06-30-interception-seams.md) 可在不改变本机制的前提下添加该门禁。
- 只读行为在 `tools/pre-execute` 拒绝门禁强制执行之前仅为建议性质；[拦截点](../../implemented/feature/2026-06-30-interception-extension-points.md) 可在不改变本机制的前提下添加该门禁。
- 经过压缩（compaction）的源会话 fork 出的是其压缩视图，因此绑定的界面应当告知用户子会话继承的是摘要而非被替换的轮次。
- 反复的 handback 会消耗父会话上下文。每次合并的长度上限约束了单条笔记的大小；后续的合并整理属于上下文压缩的职责。

4
.agents/notes/proposed/feature/2026-07-17-sdk-follow-up-capabilities.i18n.yaml
View File

@@ -2,5 +2,5 @@
# side as of the last confirmed-consistent state. Both languages carry equal authority;
# after editing either side, bring the other along and re-record with:
# pnpm run verify-translation-pairing --write .agents/notes/proposed/feature/2026-07-17-sdk-follow-up-capabilities.md
2026-07-17-sdk-follow-up-capabilities.md: 88d5d2f9bd1ce01c20177bcaee5bbe6b434bb978
2026-07-17-sdk-follow-up-capabilities.zh.md: d54dc4d0635ef37d7196ff5cecbdd6ed093c98e3
2026-07-17-sdk-follow-up-capabilities.md: 9a17f07139014f95666789e41cacb7af180ef5d8
2026-07-17-sdk-follow-up-capabilities.zh.md: 2431c5c3d9e45605223d9bb1843e48a3c711759e

2
.agents/notes/proposed/feature/2026-07-17-sdk-follow-up-capabilities.md
View File

@@ -12,7 +12,7 @@ These gaps are coupled. Create and config already share questions, feature confi
## Proposal
The SDK extends the existing prompt and project-editing boundaries instead of creating parallel workflows. A non-interactive prompt port and structured feature plan drive create and config, `dsh-sdk create <source>` delegates dependency resolution to the project package manager before mounting the resolved package through `ProjectEditSession`, launcher-side telemetry wraps `create-sdk` and every `dsh-sdk` command, and injected prompt streams provide the primary interactive-test seam.
The SDK extends the existing prompt and project-editing boundaries instead of creating parallel workflows. A non-interactive prompt port and structured feature plan drive create and config, `dsh-sdk create <source>` delegates dependency resolution to the project package manager before mounting the resolved package through `ProjectEditSession`, launcher-side telemetry wraps `create-sdk` and every `dsh-sdk` command, and injected prompt streams provide the primary interactive-test hook.
| Capability | Product entrypoint | Owning mechanism | Required outcome |
|---|---|---|---|

2
.agents/notes/proposed/feature/2026-07-17-sdk-follow-up-capabilities.zh.md
View File

@@ -12,7 +12,7 @@ Status: proposed
## 提案
SDK 扩展现有提示词与工程编辑边界，不另建平行工作流。非交互式 `PromptPort` 实现和结构化功能计划驱动 create 与 config；`dsh-sdk create <source>` 先把依赖解析交给工程的包管理器，再通过 `ProjectEditSession` 挂载解析所得的包；启动器侧遥测包住 `create-sdk` 和每个 `dsh-sdk` 命令；交互测试主要通过注入的提示词输入输出流完成。
SDK 扩展现有提示词与工程编辑边界，不另建平行工作流。非交互式 `PromptPort` 实现和结构化功能计划驱动 create 与 config；`dsh-sdk create <source>` 先把依赖解析交给工程的包管理器，再通过 `ProjectEditSession` 挂载解析所得的包；启动器侧遥测包住 `create-sdk` 和每个 `dsh-sdk` 命令；注入的提示词输入输出流提供主要的交互测试钩子。
| 功能 | 产品入口 | 所属机制 | 必须达到的结果 |
|---|---|---|---|