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

docs(notes): archive low-value decision records

Browse Source
This commit is contained in:
Tianyi Cui
2026-07-26 23:06:00 +08:00
parent 7c038f275d
commit 37140bf823
281 changed files with 1033 additions and 707 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
.agents/notes/rejected/feature/2026-07-13-stream-workflow-progress-through-tool-calls.i18n.yaml
View File

@@ -1,6 +0,0 @@
# Bilingual-pair consistency record (docs/i18n/README.md): the git blob hash of each
# 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
2026-07-13-stream-workflow-progress-through-tool-calls.md: 1b299ec323d32745cc504a90948b63a4dcaae64f
2026-07-13-stream-workflow-progress-through-tool-calls.zh.md: f84ff3bb22a19ed7ad2f9fc262a6702e254972ad

43
.agents/notes/rejected/feature/2026-07-13-stream-workflow-progress-through-tool-calls.md
View File

@@ -1,43 +0,0 @@
# Agent Note: Stream workflow progress through tool calls
Status: rejected — ACP is automation-only; live workflow presentation needs a human-interface owner and a fresh design.
English | [中文](2026-07-13-stream-workflow-progress-through-tool-calls.zh.md)
## Problem
The workflow engine intentionally emits balanced `workflow/*` observation events for run, phase, narration, and child-agent progress, but no production consumer presents them. Editors therefore show one pending workflow tool card until the final result even while the engine already reports which phase is active, what the script logged, and which children started or settled. The [dynamic-workflows decision](../../implemented/feature/2026-07-05-dynamic-workflows.md) explicitly reserves ACP progress UI for this event stream.
Making `dsh-acp` listen to workflow events directly would invert the capability boundary: the generic UI bridge would depend on an optional workflow package and special-case one tool name. The tool pipeline already owns the routing facts a live update needs—agent and call id—but exposes only pure pending/final presenters, so a long-running tool has no provider-neutral way to report transient UI state between them.
## Proposal
Add a live progress channel to `dsh-tools`. The registry-owned `ToolExecution` gains `reportProgress(view): boolean`, where `view` is a detached provider-neutral generic progress snapshot containing an optional replacement title and UI-facing content blocks. Progress cannot change the call's args-derived card tag, kind, raw input, locations, terminal intent, or diff intent; it updates only the live title/content within the presentation chosen up front. While the execution is active, the method validates and snapshots the view, then dispatches a contained, agent-scoped `tools/progress` observation carrying the authoritative execution identity and snapshot. Once final-result processing begins it returns `false` and emits nothing, so a late asynchronous reporter cannot overwrite a terminal card. Observer exceptions are logged and cannot fail the tool.
`dsh-acp` consumes `tools/progress` generically. It resolves the execution's agent through its existing agent-to-session map and emits an in-progress `tool_call_update` for the same call id. Because reporting is available only inside the tool execution pipeline, the durable `tool/call` and its ACP `tool_call` always precede the first update; closing the reporter before `tools/result` ensures no progress update follows the completed/failed card. Progress is live UI state rather than model input or durable history: session replay continues to reconstruct the pending and final cards from `tool/call` and `tool/result` without replaying transient updates.
`dsh-tool-workflow` becomes the first producer. Each tool execution installs a compact event capture before calling `ctx.workflows.start()`, because a valid engine may emit progress synchronously inside `start()`. Until the call returns, the capture reduces observed events into candidate states keyed by `WorkflowRunInfo.id`; it then selects the returned `WorkflowRun.id`, discards other candidates, reports the accumulated snapshot, and routes later matching events directly. If `start()` throws, the capture is disposed and its candidates are dropped. This preserves engine swappability without adding observer correlation to `WorkflowStartRequest` or requiring progress to wait until `start()` returns.
The reducer consumes the existing start, phase, log, agent-start, agent-end, and end events, reporting a replacement snapshot with the current phase, latest log line, active child labels, and completed/failed/cancelled counts. It does not accumulate a narration transcript; settled children leave the active set and become counters. `workflow/end`, tool settlement, or plugin disposal removes the reducer entry and event capture. The six workflow events, their metadata, paired child lifecycle, run handle, cancellation channels, and observer containment remain unchanged; third-party observers can continue consuming them directly.
Update the tool execution/presentation docs, generated event and API catalogs, workflow package docs, and the workflow data-structure catalog. ACP integration coverage must exercise the real workflow tool and worker seam with a scripted model boundary; the primary ACP snapshot suite adds one workflow-progress scenario because this changes the editor-facing transcript.
## Alternatives considered
**Delete the workflow observation surface.** Rejected in [the collapse-workflow simplification](../../rejected/simplification/2026-07-12-collapse-workflow-to-foreground-core.md): the events and their balanced lifecycle are intentional, and the missing piece is a consumer.
**Teach ACP about workflows directly.** This could map `WorkflowRunInfo` to a session and card, but it would make the generic bridge depend on an optional capability and bypass the rule that tools own presentation intent. A tool-progress channel solves the same routing problem for every long-running tool.
**Persist every progress update as a session event.** That would make live narration replayable, but it would permanently enlarge logs with state whose authoritative durable outcome is already the tool call/result pair. If resumable workflow progress becomes a product requirement, it needs a workflow-journaling design rather than UI snapshots disguised as durable facts.
## Acceptance criteria
- `ToolExecution.reportProgress()` is registry-owned, agent-scoped, snapshotting, observer-contained, and returns `false` without dispatch after terminal processing starts.
- ACP routes progress to the correct call in the correct live session; concurrent workflows in different sessions cannot cross-talk, and no `tool_call_update` appears before its `tool_call` or after its terminal update.
- Workflow progress shows the current phase, latest log line, active children, and outcome counts while preserving all existing `workflow/*` events and run semantics; a seam test engine that emits start, phase, log, child, and end events synchronously inside `start()` loses none of that reducer state.
- Cancellation, worker death, tool failure, session close, and plugin disposal release reducer state; replay emits only the durable pending/final card pair.
- Unit, workflow integration, ACP integration, snapshot, typecheck, coverage, doc-sync, module-graph, build, and hygiene gates pass.
## Risks
This adds a public live-progress method and event to the tool seam, so implementations must keep the active/terminal boundary exact and detach snapshots before observers see them. The pre-start capture can briefly observe unrelated workflow runs, so it holds only compact candidate state keyed by run id and drops every non-matching candidate as soon as `start()` returns. A workflow can emit many progress changes; the bounded reducer avoids transcript growth but still sends one UI update per meaningful event after correlation. If measured clients need coalescing, it must be a defaulted validated bridge configuration rather than a hardcoded throttle. Transient progress intentionally disappears on replay, so the final tool result remains the only durable workflow card content.

43
.agents/notes/rejected/feature/2026-07-13-stream-workflow-progress-through-tool-calls.zh.md
View File

@@ -1,43 +0,0 @@
# Agent Note: 通过工具调用流式传输工作流进度
Status: rejected — ACP 仅面向自动化；实时工作流展示需要一个面向人类界面的归属方和全新设计。
[English](2026-07-13-stream-workflow-progress-through-tool-calls.md) | 中文
## 问题
工作流引擎有意为 run、phase、narration 和子 agent（智能体）进度发出成对的 `workflow/*` observation 事件，但目前没有生产消费方呈现这些事件。因此，编辑器在最终结果返回之前只显示一张 pending 状态的工作流工具卡片，尽管引擎已经报告了当前活跃的 phase、脚本日志内容以及哪些子 agent 已启动或已结束。[动态工作流决策](../../implemented/feature/2026-07-05-dynamic-workflows.md)明确将 ACP（Agent Client Protocol）进度 UI 保留给这一事件流。
如果让 `dsh-acp` 直接监听工作流事件，就会反转能力边界：通用的 UI 桥接层将依赖一个可选的工作流包（package），并对一个工具名做特殊处理。工具流水线已经拥有实时更新所需的路由信息（agent 和 call id），但只暴露了纯粹的 pending/final 展示器，因此长时间运行的工具没有提供方无关的方式在二者之间报告瞬态 UI 状态。
## 提案
为 `dsh-tools` 添加一条实时进度通道。注册表所有的 `ToolExecution` 新增 `reportProgress(view): boolean`，其中 `view` 是一个独立的、提供方无关的通用进度快照，包含可选的替换标题和面向 UI 的内容块。进度不能更改调用的 args 派生卡片标签、kind、原始输入、locations、terminal intent 或 diff intent；它只更新在最初选定的展示方式内的实时标题/内容。当执行处于活跃状态时，该方法校验并快照 view，然后分发一个受限的、agent 作用域的 `tools/progress` observation，携带权威的执行标识与快照。一旦 final-result 处理开始，方法返回 `false` 且不再分发，因此迟到的异步报告者无法覆盖终态卡片。观察者异常会被记录日志，不会导致工具失败。
`dsh-acp` 以通用方式消费 `tools/progress`。它通过既有的 agent 到会话映射解析执行所属的 agent，并为同一 call id 发出 in-progress 的 `tool_call_update`。由于报告仅在工具执行流水线内可用，持久化的 `tool/call` 及其 ACP `tool_call` 始终先于第一条 update；在 `tools/result` 之前关闭报告者，确保进度更新不会出现在 completed/failed 卡片之后。进度是实时 UI 状态，而非模型输入或持久历史：会话回放继续从 `tool/call` 和 `tool/result` 重建 pending 与 final 卡片，无需重放瞬态更新。
`dsh-tool-workflow` 成为第一个生产者。每次工具执行在调用 `ctx.workflows.start()` 之前安装一个紧凑的事件捕获器，因为合法的引擎可能在 `start()` 内部同步发出进度。在调用返回之前，捕获器将观察到的事件按 `WorkflowRunInfo.id` 归约为候选状态；随后选取返回的 `WorkflowRun.id`，丢弃其他候选，报告累积的快照，并将后续匹配事件直接路由。如果 `start()` 抛出异常，捕获器被 dispose（资源释放），其候选状态被丢弃。这在不向 `WorkflowStartRequest` 添加观察者关联、也不要求进度等到 `start()` 返回的前提下，保持了引擎的可替换性。
归约器消费既有的 start、phase、log、agent-start、agent-end 和 end 事件，报告一个替换快照，包含当前 phase、最新日志行、活跃子 agent 标签以及 completed/failed/cancelled 计数。它不累积 narration transcript（文本记录）；已结束的子 agent 离开活跃集合，变为计数器。`workflow/end`、工具结算或插件 dispose 移除归约器条目和事件捕获器。六种工作流事件及其元数据、成对的子 agent 生命周期、run handle、取消通道和观察者隔离保持不变；第三方观察者可继续直接消费这些事件。
更新工具执行/展示文档、生成的事件与 API 目录、工作流包文档以及工作流数据结构目录。ACP 集成覆盖率必须使用脚本化的模型边界测试真实的工作流工具和 worker seam；主 ACP 快照套件新增一个 workflow-progress 场景，因为这改变了面向编辑器的 transcript。
## 曾考虑的替代方案
**删除工作流 observation 表面。** 在[折叠工作流简化提案](../../rejected/simplification/2026-07-12-collapse-workflow-to-foreground-core.md)中被否决：这些事件及其成对生命周期是有意设计的，缺少的是消费方。
**让 ACP 直接了解工作流。** 这可以将 `WorkflowRunInfo` 映射到会话和卡片，但会使通用桥接层依赖一个可选能力，并绕过「工具拥有展示意图」的规则。工具进度通道为每个长时间运行的工具解决了相同的路由问题。
**将每条进度更新持久化为会话事件。** 这会使实时 narration 可回放，但会用一种状态永久膨胀日志，而该状态的权威持久结果已经是工具调用/结果对。如果可恢复的工作流进度成为产品需求，需要一个工作流日志化设计，而非伪装成持久事实的 UI 快照。
## 验收标准
- `ToolExecution.reportProgress()` 由注册表所有、agent 作用域、快照化、观察者隔离，且在终态处理开始后返回 `false` 而不分发。
- ACP 将进度路由到正确的实时会话中的正确调用；不同会话中的并发工作流不能串扰，且 `tool_call_update` 不会出现在其 `tool_call` 之前或终态更新之后。
- 工作流进度显示当前 phase、最新日志行、活跃子 agent 和结果计数，同时保留所有既有 `workflow/*` 事件和 run 语义；一个在 `start()` 内部同步发出 start、phase、log、child 和 end 事件的 seam 测试引擎不会丢失任何归约器状态。
- 取消、worker 死亡、工具失败、会话关闭和插件 dispose 释放归约器状态；回放仅发出持久的 pending/final 卡片对。
- 单元测试、工作流集成测试、ACP 集成测试、快照、类型检查、覆盖率、doc-sync（文档同步门禁）、module-graph、构建和 hygiene 门禁全部通过。
## 风险
本提案向工具 seam 添加了一个公开的实时进度方法和事件，因此实现方必须精确维护 active/terminal 边界，并在观察者看到快照之前将其分离。pre-start 捕获器可能短暂观察到无关的工作流 run，因此它仅按 run id 持有紧凑的候选状态，并在 `start()` 返回后立即丢弃所有不匹配的候选。一个工作流可能发出大量进度变更；有界归约器避免了 transcript 增长，但在关联完成后仍会为每个有意义的事件发送一条 UI 更新。如果经测量的客户端需要合并更新，这必须是一个带默认值的、经过校验的桥接配置，而非硬编码的节流。瞬态进度在回放时有意消失，因此最终工具结果仍是唯一持久的工作流卡片内容。