feat: add persistent PTY sessions

.agents/notes/implemented/feature/2026-07-16-persistent-pty-sessions.i18n.yaml

@@ -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
-2026-07-16-persistent-pty-sessions.md: 3028d3a527e177f2b30c557dc45443af99783d6c
-2026-07-16-persistent-pty-sessions.zh.md: f244992abccc9c107bc2cf4da392dc39d39d14cc
+2026-07-16-persistent-pty-sessions.md: ef2d149c9d7ba4b1df03f003166f94a2996e5d45
+2026-07-16-persistent-pty-sessions.zh.md: 7f2ee00804b6971b5c64242c8ee84d1631e8b656

.agents/notes/implemented/feature/2026-07-16-persistent-pty-sessions.md

@@ -1,6 +1,6 @@
 # Agent Note: persistent PTY sessions
 
-Status: proposed
+Status: implemented
 
 English | [中文](2026-07-16-persistent-pty-sessions.zh.md)
 
@@ -12,11 +12,11 @@ That gap excludes workflows whose state lives in a terminal rather than a file:
 
 The existing `bash`, `read`, `write`, and `edit` tools remain the reliable default for bounded, auditable operations. A PTY is an additional capability for work that genuinely requires terminal state, not evidence that those tools are defective or candidates for removal.
 
-## Proposal
+## Decision
 
-Add an optional `packages/pty/` capability family that exposes agent-owned, persistent, line-oriented PTY sessions. It follows the repository's [capability pattern](../../implemented/architecture/2026-06-13-capability-seams.md), coexists with the existing command and filesystem tools, and does not change `agent-loop`.
+The optional `packages/pty/` capability family exposes agent-owned, persistent, line-oriented PTY sessions. It follows the repository's [capability pattern](../../implemented/architecture/2026-06-13-capability-seams.md), coexists with the existing command and filesystem tools, and does not change `agent-loop`.
 
-The first delivery supports interactive shells and line-oriented REPLs on Linux and macOS. Full-screen terminal applications, keystroke sequences, BEL-triggered control flow, session restoration after process loss, and cross-agent session sharing are explicitly deferred until the basic lifecycle is proven.
+The implementation supports interactive shells and line-oriented REPLs on Linux and macOS. Full-screen terminal applications, keystroke sequences, BEL-triggered control flow, session restoration after process loss, and cross-agent session sharing are explicitly deferred.
 
 ### Package topology
 
@@ -32,7 +32,7 @@ Idle detection is backend behavior, not a second public seam. A remote or contai
 
 `PtyService` stores live sessions process-locally, but every session is owned by the exact `Agent` passed through the tool execution context. The service mints an opaque `PtySessionId`; an optional model-chosen `name` is display metadata and is unique only within that owner. Every operation targets `sessionId`, and `list`/`read`/`signal`/`kill` reject callers other than the owner.
 
-The initial design has no plugin-load auto-start sessions. `pty_spawn` creates a session only during an agent tool call, when ownership and the owning event-sourced session are known. Deployments that later need declarative startup must compose it through unpublished agent setup rather than create shared global terminals.
+There are no plugin-load auto-start sessions. `pty_spawn` creates a session only during an agent tool call, when ownership and the owning event-sourced session are known. A future declarative startup feature must compose through unpublished agent setup rather than create shared global terminals.
 
 Agent-scope disposal closes registrations first, then awaits quiescent teardown of every owned PTY. Backend or tool-plugin reload does not orphan sessions: ownership lives in `PtyService` until the agent ends, following the same service-owned-record pattern as [`ctx.tasks`](../../../../packages/tasks/tasks/README.md).
 
@@ -41,11 +41,11 @@ Agent-scope disposal closes registrations first, then awaits quiescent teardown
 A registered `shell` backend constrains how a terminal starts; it does not constrain commands typed after startup. `dsh-pty-local` therefore applies two protections before spawning:
 
 - It builds a scrubbed child environment using the same credential-shaped-name policy as `bash-local`, removing ambient `*KEY*`, `*SECRET*`, `*TOKEN*`, and harness-managed variables unless an explicit trusted mapping supplies them.
-- Its `sandbox` config is `required | optional | disabled`, defaulting to `required`. `required` fails plugin load when `ctx.sandbox` is unavailable; `optional` uses the provider when present; `disabled` is an explicit unconfined opt-in. The selected provider wraps the session argv once and remains the process boundary for the PTY lifetime.
+- It requires `ctx.sandbox` and the shared `ctx.sandboxPolicy`. At spawn, the backend resolves the owner's effective session mode over the deployment default and wraps the shell argv once; that mode and workspace root remain the process boundary for the PTY lifetime. `danger-full-access` is the existing explicit unconfined choice rather than a PTY-specific bypass.
 
 Sandboxing confines local process effects but does not make arbitrary shell input safe: network calls and other external side effects remain governed by deployment policy. Tool descriptions state that PTY sessions are less auditable than one-shot tools and should be used only when persistence or interactive stdin is necessary.
 
-The implementation uses only public `node-pty` capabilities: child PID, `data` and `exit` notifications, `write`, `resize`, and `kill`. It does not assume access to the native master fd or call `waitpid` from TypeScript. Platform process inspectors derive process-group and session membership from `/proc` on Linux and `ps` on macOS.
+The implementation uses only public `node-pty` capabilities: child PID, `data` and `exit` notifications, `write`, `resize`, and `kill`. It does not assume access to the native master fd or call `waitpid` from TypeScript. Platform process inspectors derive foreground process groups and parent/child identity from `/proc` on Linux and `ps` on macOS.
 
 ### Six model-facing tools
 
@@ -58,11 +58,11 @@ The implementation uses only public `node-pty` capabilities: child PID, `data` a
 | `pty_kill` | Close one session and await process-tree quiescence | `{ killed }` |
 | `pty_list` | List the caller's live sessions | owner-scoped session summaries |
 
-`pty_send({ sessionId, text, submit?, background? })` treats `text` as UTF-8 bytes and resolves `submit` to `true` in the tool implementation. When `submit` is true it writes the platform Enter sequence after the text; when false it writes only the text, allowing control characters and REPL fragments without hidden content heuristics.
+`pty_send({ sessionId, text, submit?, run_in_background? })` treats `text` as UTF-8 bytes and resolves `submit` to `true` in the tool implementation. When `submit` is true it writes the platform Enter sequence after the text; when false it writes only the text, allowing control characters and REPL fragments without hidden content heuristics.
 
 Foreground sends return a bounded rendered delta and two independent facts: `waitReason` (`stdin_read | inferred_idle | timeout | session_exit`) and `sessionStatus` (`running` or `exited` with exit code or signal). `session_exit` refers to the PTY's top-level shell process, not an arbitrary foreground command whose status the shell consumes. A timeout never implies process exit.
 
-With `background: true`, `dsh-tool-pty` registers the in-flight send on `ctx.tasks` and returns immediately with `taskId`. `task_output(wait: true)` waits, reads incremental output, and records the final result; `task_kill` forwards cancellation as `SIGINT` and escalates only through the PTY backend's owned teardown path. If the task surface is absent, background mode fails before writing input. No PTY-specific `sleep` tool or general wake-up seam is added.
+With `run_in_background: true`, `dsh-tool-pty` registers the in-flight send on `ctx.tasks` and returns immediately with `taskId`. `task_output(wait: true)` waits, reads incremental output, and records the final result; `task_kill` forwards cancellation as `SIGINT` and escalates only through the PTY backend's owned teardown path. If the task surface is absent, background mode fails before writing input. No PTY-specific `sleep` tool or general wake-up seam is added.
 
 `pty_read` pages backward from the newest retained line. The backend enforces both line and UTF-8 byte caps on retained scrollback and the complete returned value, so one oversized line cannot bypass the bound. `truncated` distinguishes retention loss from an ordinary viewport delta.
 
@@ -70,7 +70,7 @@ With `background: true`, `dsh-tool-pty` registers the in-flight send on `ctx.tas
 
 ### Local readiness detection
 
-The local backend runs three bounded tiers. All timings are validated config fields: `pollIntervalMs`, `exactProbeAfterMs`, `idleSilenceMs`, and `timeoutMs`.
+The local backend first recognizes a private OSC prompt marker emitted by its controlled bash startup, then runs three bounded fallback tiers. The marker is removed before output reaches the model and avoids a fixed silence delay for ordinary shell commands on both platforms. All timings are validated config fields: `pollIntervalMs`, `exactProbeAfterMs`, `idleSilenceMs`, and `timeoutMs`.
 
 On Linux, the inspector reads the shell's terminal foreground PGID from `/proc/<shellPid>/stat`, enumerates every process and thread in that process group, and probes their current syscalls. A positive Tier 1 result requires an observed stdin wait: direct `read(0)`, a permitted read of a `select`/`pselect6` or `poll`/`ppoll` argument containing fd 0, or an epoll interest list containing fd 0. Unreadable process memory and unrecognized syscalls are misses, never positive guesses. Architecture tables contain only syscall numbers defined by the corresponding Linux UAPI; unsupported architectures skip Tier 1.
 
@@ -78,7 +78,7 @@ On macOS there is no exact syscall tier. Output silence returns `inferred_idle`
 
 Tier 2 returns `inferred_idle` after `idleSilenceMs` without output. A sleeping or network-blocked command can therefore look ready. Tier 3 returns `timeout` after `timeoutMs` so a foreground tool call cannot hold the agent indefinitely. The result preserves the distinction; callers may wait through `ctx.tasks`, signal the foreground group, or inspect from another session.
 
-`node-pty` data notifications feed one streaming decoder and terminal parser. Parser carry state handles UTF-8 and terminal query sequences split across chunks. The first delivery normalizes line-oriented output and detects alternate-screen entry, but it does not promise correct interaction with a full-screen application.
+`node-pty` data notifications feed one streaming decoder and terminal parser. Parser carry state handles UTF-8 and terminal query sequences split across chunks. The implementation normalizes line-oriented output and detects alternate-screen entry, but it does not promise correct interaction with a full-screen application.
 
 ### Model-visible output and durability
 
@@ -88,9 +88,9 @@ Background sends use the existing task completion notice and `task_output` resul
 
 ### Process-tree teardown
 
-The top-level `node-pty` child is treated as the POSIX session leader, but the owned resource is the complete OS process session, not one PID. On close, the backend stops callbacks, sends `SIGTERM` to all still-matching session members, closes the PTY, awaits `node-pty` exit plus process-inspector quiescence, then sends `SIGKILL` to verified survivors after configurable `disposeGraceMs`. Membership snapshots include process-start identity so PID reuse cannot redirect escalation.
+The top-level `node-pty` child is the ownership anchor. On close, the backend stops callbacks, snapshots that PID and its transitive descendants by parent PID in children-first order, sends `SIGTERM`, closes the PTY, waits for quiescence, then sends `SIGKILL` to verified survivors after configurable `disposeGraceMs` and waits for them to leave the process table. Every captured PID includes process-start identity so reuse cannot redirect escalation.
 
-Teardown reports root exit and survivor cleanup independently. It does not claim success merely because the shell exited; disposal resolves only after no captured session member remains or returns a structured cleanup failure naming the survivors.
+Teardown reports root exit and survivor cleanup independently. It does not claim success merely because the shell exited; disposal resolves only after no captured tree member remains or returns a structured cleanup failure naming the survivors. It never broadens ownership to every member of the root PID's POSIX session.
 
 ### Composition and rollout
 
@@ -99,10 +99,13 @@ The example composition remains opt-in and safe by default:
 ```yaml
 plugins:
   '@deepseek-ai/dsh-sandbox-local':
+  '@deepseek-ai/dsh-sandbox-policy':
+    config:
+      mode: workspace-write
+      workspaceRoot: .
   '@deepseek-ai/dsh-pty':
   '@deepseek-ai/dsh-pty-local':
     config:
-      sandbox: required
       scrollbackLines: 10000
       scrollbackMaxBytes: 4194304
       maxReadBytes: 262144
@@ -114,7 +117,7 @@ plugins:
   '@deepseek-ai/dsh-tool-pty':
 ```
 
-The package ships concise tool guidance explaining persistent state, owner isolation, uncertain idle results, cleanup, and the preference for existing one-shot tools when interaction is unnecessary. It does not add a global system-prompt recommendation or mount PTY in shipped defaults.
+The package ships concise tool guidance explaining persistent state, owner isolation, uncertain idle results, cleanup, and the preference for existing one-shot tools when interaction is unnecessary. It does not add a global system-prompt recommendation or mount PTY in shipped defaults; dedicated ACP and headless snapshot overlays exercise the opt-in composition.
 
 ### Deferred work
 
@@ -130,39 +133,37 @@ The package ships concise tool guidance explaining persistent state, owner isola
 
 **Add persistent mode to `bash`.** Rejected. Returning on readiness rather than process exit, retaining a process tree across calls, and exposing interactive stdin create a different ownership and failure contract.
 
-**Require native master-fd access from `node-pty`.** Rejected. Its public API exposes no master fd. The local backend instead derives foreground and session membership from supported OS process metadata and treats unreadable metadata as a detector miss.
+**Require native master-fd access from `node-pty`.** Rejected. Its public API exposes no master fd. The local backend instead derives foreground groups and descendants from supported OS process metadata and treats unreadable metadata as a detector miss.
+
+**Signal every member of the root PID's POSIX session.** Rejected. `node-pty` may expose a helper PID whose session belongs to the launcher, so SID-wide teardown can signal unrelated harness or desktop processes. A PID-identity-fenced descendant tree is narrower and safe by construction.
 
 **Publish `PtyIdleDetector` as a replaceable registry.** Rejected. Only the local backend needs these platform probes, while remote backends may receive readiness over their own protocol. Backend replacement already provides the necessary extension point.
 
 **Add a PTY-specific `sleep` tool.** Rejected. `ctx.tasks` already owns bounded waiting, cancellation, completion notices, and model-facing collection. A second general wake mechanism would cross the agent-loop boundary and duplicate that contract.
 
-**Include TUI sequences and BEL handling in the first delivery.** Rejected. The source prototype treats those paths as timing-sensitive and still records unresolved alternate-screen and interaction failures. Line-oriented PTY use proves the core value without making those unverified behaviors foundational.
+**Include TUI sequences and BEL handling.** Rejected. The source prototype treats those paths as timing-sensitive and still records unresolved alternate-screen and interaction failures. Line-oriented PTY use proves the core value without making those unverified behaviors foundational.
 
 **Use an out-of-process daemon immediately.** Rejected for the initial in-process capability because current persistent front doors already keep a Cordis context alive. A daemon becomes justified by cross-process restoration or multi-client attachment, both deferred here.
 
-## Acceptance criteria
+## Verification
 
-- `packages/pty/{pty,pty-local,tool-pty}` build as the interface, local implementation, and model consumer; backend registrations dispose cleanly.
-- Every live PTY has one service-minted `PtySessionId`, one exact `Agent` owner, owner-fenced operations, and awaited cleanup on agent disposal; concurrent agents may reuse display names without sharing state.
-- `dsh-pty-local` uses only public `node-pty` APIs and contains no master-fd or TypeScript `waitpid` assumption.
-- Environment tests prove credential-shaped ambient variables are absent. `sandbox: required` fails at load without a provider, and real composition proves the provider wraps the long-lived session process.
-- Linux fixtures cover pipelines, a stdin-reading non-leader process, a stdin-reading non-main thread, unreadable process memory, supported UAPI syscall tables, unsupported architectures, and false-positive rejection. macOS process-inspector logic reaches 100% coverage on Linux, and macOS CI drives a real bash and Python REPL.
-- Foreground tests exercise `stdin_read`, `inferred_idle`, `timeout`, and top-level session exit without treating a foreground command exit as directly observable.
-- Background sends register `ctx.tasks` work, return before readiness, stream bounded output through `task_output`, honor task cancellation, and fail before writing when the task surface is absent.
-- Scrollback and every model-facing result enforce final UTF-8 byte bounds, including a single oversized line and multibyte boundary cases.
-- `pty_signal` resolves the live foreground group, rejects lookup failure and shell-targeted `SIGKILL`, and never falls back to a guessed PID.
-- Disposal tests start foreground and background descendants, including a signal-ignoring child, then prove every captured process identity is gone immediately after awaited agent disposal.
-- A test-only `cordis.yml` boots through the Loader on Linux and macOS, mounts the real local backend plus sandbox, and drives spawn/send/read/signal/kill/list through the real tool registry. ACP and headless snapshots pin the six schemas, bounded results, errors, and render intents.
-- TUI, sequence, BEL, auto-start, Windows, and crash-restoration behavior are absent from the public schema and documented as deferred rather than simulated by fixtures.
-- Package READMEs and JSDoc document configuration, ownership, failure, cancellation, bounds, sandboxing, model-visible effects, and limitations; `docs/architecture.md` and generated catalogs update with the implementation.
-- The repository CI-equivalent sequence in root `AGENTS.md` passes, including `test:coverage`, snapshots, documentation, build, hygiene, and built-entry smokes.
+- Per-file coverage pins owner fencing, concurrent reservations, lifecycle cleanup, readiness tiers, sanitizer carry state, UTF-8 bounds, task integration, schemas, and render intents.
+- Linux process fixtures cover non-leader and non-main-thread stdin waits, unreadable process state, supported syscall tables, unsupported architectures, and false-positive rejection; macOS inspector logic is injected into the same unit suite.
+- Real `node-pty` tests exercise shell state, shared sandbox policy, environment scrubbing, signals, a TERM-ignoring descendant, and immediate post-disposal quiescence on supported hosts.
+- A Loader-driven `cordis.yml` test mounts the real three-package composition, while ACP and headless snapshots pin the six schemas, bounded results, error rendering, and terminal/generic cards through opt-in overlays.
+- Package contracts, the architecture map, core data structures, generated catalogs, and the website API describe the same shipped surface.
+- The repository CI-equivalent sequence owns type, lint, coverage, snapshot, documentation, build, hygiene, demo, and built-entry verification.
 
-## Risks
+## Consequences
+
+**Persistent terminal state is available without weakening one-shot tools.** Shell and REPL state can survive tool calls, while `bash`, `read`, `write`, and `edit` retain their narrower validation, approval, and replay contracts.
 
 **Idle below Linux Tier 1 is heuristic.** Output silence cannot distinguish a prompt from sleep or network I/O. The typed result preserves uncertainty, and bounded timeout plus task waiting and signaling keep control with the model.
 
 **Persistent state can drift from the model's belief.** The model may forget its cwd or active REPL. Session summaries and retained output help recovery, but no prompt can make state persistence deterministic.
 
+**A daemonized descendant can leave the captured tree.** A process that reparents before teardown is no longer discoverable from the `node-pty` root. The implementation accepts that cleanup gap instead of risking SID-wide signals to unrelated processes.
+
 **A shell can cause external side effects.** Session sandboxing and environment scrubbing reduce local exposure but do not undo pushes, API calls, or messages. Deployments that cannot tolerate those effects must omit PTY or add network policy.
 
 **Process loss destroys terminal state.** In-process sessions do not survive a harness crash or restart, and raw scrollback is not durable. Important work must be committed to files or another durable system.

.agents/notes/implemented/feature/2026-07-16-persistent-pty-sessions.zh.md

@@ -1,6 +1,6 @@
 # Agent Note: 持久化 PTY 会话
 
-Status: proposed
+Status: implemented
 
 [English](2026-07-16-persistent-pty-sessions.md) | 中文
 
@@ -12,11 +12,11 @@ harness 可以运行前台与后台命令、编辑文件和委派工作，但无
 
 现有 `bash`、`read`、`write` 和 `edit` 工具仍是有界、可审计操作的可靠默认选项。PTY 是对确实需要终端状态的工作的补充功能，不说明这些工具有缺陷，更不意味着要移除它们。
 
-## 提案
+## 决策
 
-新增可选的 `packages/pty/` 功能家族，向模型提供由 agent 拥有、持久化且面向行式交互的 PTY 会话。它遵循仓库的 [capability pattern](../../implemented/architecture/2026-06-13-capability-seams.md)，与现有命令和文件系统工具并存，并且不修改 `agent-loop`。
+可选的 `packages/pty/` 功能家族提供由 agent 拥有、持久化且面向行式交互的 PTY 会话。它遵循仓库的 [capability pattern](../../implemented/architecture/2026-06-13-capability-seams.md)，与现有命令和文件系统工具并存，并且不修改 `agent-loop`。
 
-首次交付在 Linux 和 macOS 上支持交互式 shell 与行式 REPL。全屏终端应用、按键序列、BEL 触发的控制流、进程丢失后的会话恢复以及跨 agent 共享会话都明确推迟，直到基础生命周期得到验证。
+当前实现在 Linux 和 macOS 上支持交互式 shell 与行式 REPL。全屏终端应用、按键序列、BEL 触发的控制流、进程丢失后的会话恢复以及跨 agent 共享会话都明确推迟。
 
 ### 包拓扑
 
@@ -32,7 +32,7 @@ idle 检测属于后端行为，不是第二条公共 seam。远程或容器后
 
 `PtyService` 在进程内保存活会话，但每个会话都由工具执行上下文传入的确切 `Agent` 拥有。服务铸造不透明的 `PtySessionId`；模型可选填的 `name` 只是显示元数据，仅在该 owner 内唯一。所有操作都以 `sessionId` 为目标，`list`/`read`/`signal`/`kill` 会拒绝 owner 之外的调用方。
 
-初始设计不提供插件加载期 auto-start 会话。`pty_spawn` 只在 agent 工具调用期间创建会话，此时所有权和所属的事件溯源会话都已确定。若部署后续需要声明式启动，必须通过尚未发布的 agent setup 组合，而不能创建全局共享终端。
+实现不提供插件加载期 auto-start 会话。`pty_spawn` 只在 agent 工具调用期间创建会话，此时所有权和所属的事件溯源会话都已确定。未来的声明式启动功能必须通过尚未发布的 agent setup 组合，而不能创建全局共享终端。
 
 agent scope dispose 时先关闭注册，再等待全部所属 PTY 静默退出。后端或工具插件 reload 不会遗留会话：所有权持续存放在 `PtyService` 中，直到 agent 结束，与 [`ctx.tasks`](../../../../packages/tasks/tasks/README.md) 的服务持有记录模式一致。
 
@@ -41,11 +41,11 @@ agent scope dispose 时先关闭注册，再等待全部所属 PTY 静默退出
 注册的 `shell` 后端只约束终端如何启动，不约束启动后输入的命令。因此 `dsh-pty-local` 在 spawn 前应用两层保护：
 
 - 它使用与 `bash-local` 相同的凭证形态名称策略构建清洗后的子进程环境，移除环境中的 `*KEY*`、`*SECRET*`、`*TOKEN*` 和 harness 管理的变量，除非显式的可信映射提供这些值。
-- 它的 `sandbox` 配置为 `required | optional | disabled`，默认 `required`。`required` 在缺少 `ctx.sandbox` 时于插件加载期失败；`optional` 在提供方存在时使用；`disabled` 是显式选择无约束模式。所选提供方只包装一次会话 argv，并在 PTY 的整个生命周期中充当进程边界。
+- 它要求 `ctx.sandbox` 和共享的 `ctx.sandboxPolicy`。后端在 spawn 时，以部署默认值为底折叠 owner 的有效 session mode，并只包装一次 shell argv；该 mode 与 workspace root 在 PTY 的整个生命周期中充当进程边界。`danger-full-access` 是现有的显式无约束选择，不另设 PTY 私有 bypass。
 
 沙箱限制本地进程副作用，但不会让任意 shell 输入自动安全：网络调用和其他外部副作用仍由部署策略治理。工具描述会说明 PTY 会话比一次性工具更难审计，只应在确实需要持久状态或交互式 stdin 时使用。
 
-实现只使用 `node-pty` 的公共功能：子进程 PID、`data` 与 `exit` 通知、`write`、`resize` 和 `kill`。它不假设能访问原生 master fd，也不从 TypeScript 调用 `waitpid`。平台进程检查器在 Linux 上通过 `/proc`、在 macOS 上通过 `ps` 推导进程组和会话成员关系。
+实现只使用 `node-pty` 的公共功能：子进程 PID、`data` 与 `exit` 通知、`write`、`resize` 和 `kill`。它不假设能访问原生 master fd，也不从 TypeScript 调用 `waitpid`。平台进程检查器在 Linux 上通过 `/proc`、在 macOS 上通过 `ps` 推导前台进程组和父子进程身份。
 
 ### 6 个面向模型的工具
 
@@ -58,11 +58,11 @@ agent scope dispose 时先关闭注册，再等待全部所属 PTY 静默退出
 | `pty_kill` | 关闭一个会话并等待进程树静默退出 | `{ killed }` |
 | `pty_list` | 列出调用方的活会话 | 按 owner 隔离的会话摘要 |
 
-`pty_send({ sessionId, text, submit?, background? })` 将 `text` 视为 UTF-8 字节，并由工具实现在解析阶段把 `submit` 默认成 `true`。`submit` 为 true 时先写入文本，再写入平台 Enter 序列；为 false 时只写文本，使控制字符和 REPL 片段无需隐藏的内容启发式即可发送。
+`pty_send({ sessionId, text, submit?, run_in_background? })` 将 `text` 视为 UTF-8 字节，并由工具实现在解析阶段把 `submit` 默认成 `true`。`submit` 为 true 时先写入文本，再写入平台 Enter 序列；为 false 时只写文本，使控制字符和 REPL 片段无需隐藏的内容启发式即可发送。
 
 前台发送返回有界的渲染增量和两个独立事实：`waitReason`（`stdin_read | inferred_idle | timeout | session_exit`）与 `sessionStatus`（`running`，或携带退出码或信号的 `exited`）。`session_exit` 指 PTY 顶层 shell 进程退出，不指由 shell 消费状态的任意前台命令。timeout 从不意味着进程已经退出。
 
-当 `background: true` 时，`dsh-tool-pty` 在 `ctx.tasks` 上注册进行中的发送，并立即返回 `taskId`。`task_output(wait: true)` 负责等待、读取增量输出并记录最终结果；`task_kill` 将取消转发为 `SIGINT`，只有 PTY 后端拥有的 teardown 路径可以升级信号。若 task 对外接口不存在，后台模式必须在写入输入前失败。设计不新增 PTY 专用的 `sleep` 工具或通用唤醒 seam。
+当 `run_in_background: true` 时，`dsh-tool-pty` 在 `ctx.tasks` 上注册进行中的发送，并立即返回 `taskId`。`task_output(wait: true)` 负责等待、读取增量输出并记录最终结果；`task_kill` 将取消转发为 `SIGINT`，只有 PTY 后端拥有的 teardown 路径可以升级信号。若 task 对外接口不存在，后台模式必须在写入输入前失败。设计不新增 PTY 专用的 `sleep` 工具或通用唤醒 seam。
 
 `pty_read` 从最新保留行向后分页。后端同时对保留的 scrollback 和完整返回值执行行数与 UTF-8 字节上限，因此单个超长行无法绕过限制。`truncated` 用于区分保留数据丢失与普通 viewport 增量。
 
@@ -70,7 +70,7 @@ agent scope dispose 时先关闭注册，再等待全部所属 PTY 静默退出
 
 ### 本地就绪检测
 
-本地后端执行 3 个有界层级。所有时间参数都是经校验的配置字段：`pollIntervalMs`、`exactProbeAfterMs`、`idleSilenceMs` 和 `timeoutMs`。
+本地后端先识别受控 bash 启动时发出的私有 OSC prompt marker，再执行 3 个有界 fallback 层级。marker 在输出到达模型前被移除，使两个平台上的普通 shell 命令都无需固定等待静默阈值。所有时间参数都是经校验的配置字段：`pollIntervalMs`、`exactProbeAfterMs`、`idleSilenceMs` 和 `timeoutMs`。
 
 在 Linux 上，检查器从 `/proc/<shellPid>/stat` 读取 shell 的终端前台 PGID，枚举该进程组中的每个进程与线程，并检查它们当前的 syscall。Tier 1 只有观察到 stdin 等待才返回正结果：直接 `read(0)`、获准读取且含 fd 0 的 `select`/`pselect6` 或 `poll`/`ppoll` 参数，或者含 fd 0 的 epoll interest list。无法读取的进程内存和未识别的 syscall 都是 miss，绝不作为正向猜测。架构表只包含对应 Linux UAPI 定义的 syscall number；不支持的架构跳过 Tier 1。
 
@@ -78,7 +78,7 @@ macOS 没有精确 syscall 层。任何前台进程组输出静默都会返回 `
 
 Tier 2 在持续 `idleSilenceMs` 没有输出后返回 `inferred_idle`，因此 sleep 或网络阻塞的命令可能看似 ready。Tier 3 在 `timeoutMs` 后返回 `timeout`，避免前台工具调用无限占住 agent。结果保留这些区别；调用方可以通过 `ctx.tasks` 等待、向前台组发信号，或从另一个会话排查。
 
-`node-pty` data 通知进入同一个流式 decoder 和终端 parser。parser 的 carry 状态处理跨 chunk 的 UTF-8 与终端查询序列。首次交付只规范化行式输出并检测 alternate-screen 进入，不承诺正确操作全屏应用。
+`node-pty` data 通知进入同一个流式 decoder 和终端 parser。parser 的 carry 状态处理跨 chunk 的 UTF-8 与终端查询序列。当前实现只规范化行式输出并检测 alternate-screen 进入，不承诺正确操作全屏应用。
 
 ### 模型可见输出与持久性
 
@@ -88,9 +88,9 @@ Tier 2 在持续 `idleSilenceMs` 没有输出后返回 `inferred_idle`，因此
 
 ### 进程树 teardown
 
-顶层 `node-pty` 子进程视为 POSIX 会话 leader，但所属资源是完整的 OS 进程会话，而不是一个 PID。关闭时，后端先停止 callback，再向仍匹配的会话成员发送 `SIGTERM`、关闭 PTY、等待 `node-pty` exit 与进程检查器确认静默，然后在可配置的 `disposeGraceMs` 后向已验证的存活者发送 `SIGKILL`。成员快照包含进程启动身份，避免 PID 复用把升级信号发给无关进程。
+顶层 `node-pty` 子进程是所有权锚点。关闭时，后端先停止 callback，再按父 PID 以子进程优先顺序捕获该 PID 及其传递子进程、发送 `SIGTERM`、关闭 PTY 并等待静默，然后在可配置的 `disposeGraceMs` 后向已验证的存活者发送 `SIGKILL`，并等待它们离开进程表。每个捕获的 PID 都包含进程启动身份，避免 PID 复用把升级信号发给无关进程。
 
-teardown 独立报告根进程退出与存活进程清理。它不会只因 shell 退出就声称成功；dispose 只有在已捕获的会话成员全部消失后才完成，否则返回结构化清理失败并列出存活者。
+teardown 独立报告根进程退出与存活进程清理。它不会只因 shell 退出就声称成功；dispose 只有在已捕获的进程树成员全部消失后才完成，否则返回结构化清理失败并列出存活者。所有权绝不会扩大到根 PID 所属 POSIX 会话的全部成员。
 
 ### 组合与推行
 
@@ -99,10 +99,13 @@ teardown 独立报告根进程退出与存活进程清理。它不会只因 shel
 ```yaml
 plugins:
   '@deepseek-ai/dsh-sandbox-local':
+  '@deepseek-ai/dsh-sandbox-policy':
+    config:
+      mode: workspace-write
+      workspaceRoot: .
   '@deepseek-ai/dsh-pty':
   '@deepseek-ai/dsh-pty-local':
     config:
-      sandbox: required
       scrollbackLines: 10000
       scrollbackMaxBytes: 4194304
       maxReadBytes: 262144
@@ -114,7 +117,7 @@ plugins:
   '@deepseek-ai/dsh-tool-pty':
 ```
 
-包会提供简洁的工具指引，说明持久状态、owner 隔离、不确定的 idle 结果、清理，以及无需交互时优先使用现有一次性工具。它不增加全局 system prompt 推荐，也不在已发布的默认配置中挂载 PTY。
+包提供简洁的工具指引，说明持久状态、owner 隔离、不确定的 idle 结果、清理，以及无需交互时优先使用现有一次性工具。它不增加全局 system prompt 推荐，也不在已发布的默认配置中挂载 PTY；专用 ACP 与 headless 快照 overlay 覆盖 opt-in 组合。
 
 ### 推迟的工作
 
@@ -130,39 +133,37 @@ plugins:
 
 **给 `bash` 增加持久模式。**拒绝。按就绪而不是进程退出返回、跨调用保留进程树、暴露交互式 stdin 会形成不同的所有权和失败契约。
 
-**要求从 `node-pty` 获取原生 master fd。**拒绝。它的公共 API 不暴露 master fd。本地后端改为从受支持的 OS 进程元数据推导前台组和 session 成员，并把不可读元数据视为 detector miss。
+**要求从 `node-pty` 获取原生 master fd。**拒绝。它的公共 API 不暴露 master fd。本地后端改为从受支持的 OS 进程元数据推导前台组与子孙进程，并把不可读元数据视为 detector miss。
+
+**向根 PID 所属 POSIX 会话的全部成员发送信号。**拒绝。`node-pty` 可能暴露属于启动器会话的 helper PID，因此按 SID 清理可能向无关的 harness 或桌面进程发送信号。带 PID 启动身份校验的子孙进程树范围更窄，其安全边界由结构保证。
 
 **发布可替换注册表 `PtyIdleDetector`。**拒绝。只有本地后端需要这些平台 probe，远程后端可能通过自己的协议接收就绪状态。替换后端已经提供所需扩展点。
 
 **新增 PTY 专用 `sleep` 工具。**拒绝。`ctx.tasks` 已经拥有有界等待、取消、完成通知和面向模型的收集。第二套通用唤醒机制会跨越 agent loop（智能体循环）边界并重复该契约。
 
-**在首次交付包含 TUI sequence 与 BEL 处理。**拒绝。源 prototype 将这些路径视为 timing-sensitive，且仍记录未解决的 alternate-screen 和交互失败。行式 PTY 已能证明核心价值，无需把未经验证的行为放进基础层。
+**包含 TUI sequence 与 BEL 处理。**拒绝。源 prototype 将这些路径视为 timing-sensitive，且仍记录未解决的 alternate-screen 和交互失败。行式 PTY 已能证明核心价值，无需把未经验证的行为放进基础层。
 
 **立即采用进程外 daemon。**初始的进程内功能不采用，因为当前持久 front door 已能维持 Cordis context。跨进程恢复或多客户端 attach 会让 daemon 变得合理，但两者都已推迟。
 
-## 验收标准
+## 验证
 
-- `packages/pty/{pty,pty-local,tool-pty}` 分别作为接口、本地实现和模型消费方构建；后端注册可干净 dispose。
-- 每个活 PTY 都有一个由服务铸造的 `PtySessionId`、一个确切的 `Agent` owner、按 owner 隔离的操作，并在 agent dispose 时等待清理；并发 agent 可以复用显示名称而不共享状态。
-- `dsh-pty-local` 只使用 `node-pty` 公共 API，不包含 master-fd 或 TypeScript `waitpid` 假设。
-- 环境测试证明凭证形态的环境变量不存在。缺少提供方时 `sandbox: required` 在加载期失败，REAL-composition 测试证明提供方包装长活会话进程。
-- Linux fixture（测试前置数据）覆盖 shell 管道、读取 stdin 的非 leader 进程、读取 stdin 的非主线程、不可读进程内存、受支持的 UAPI syscall 表、不支持的架构和误报拒绝。macOS 进程检查逻辑在 Linux 上达到 100% 覆盖率，macOS CI 驱动真实 bash 与 Python REPL。
-- 前台测试覆盖 `stdin_read`、`inferred_idle`、`timeout` 和顶层会话退出，不把前台命令退出当作可直接观察事件。
-- 后台发送注册 `ctx.tasks` work、在就绪前返回、通过 `task_output` 流式提供有界输出、遵守 task cancellation，并在 task 对外接口缺失时于写入前失败。
-- scrollback 与每个面向模型的结果都对最终 UTF-8 字节执行上限，包括单个超长行和多字节边界情况。
-- `pty_signal` 解析活跃前台组，拒绝查询失败和指向 shell 的 `SIGKILL`，且绝不回退到猜测的 PID。
-- dispose 测试启动前台与后台子进程，包括忽略信号的子进程，然后证明等待 agent dispose 后每个捕获的进程身份立即消失。
-- 测试专用 `cordis.yml` 在 Linux 与 macOS 上通过 Loader 启动，挂载真实本地后端与沙箱，并通过真实工具注册表驱动 spawn/send/read/signal/kill/list。ACP 与 headless 快照固定 6 个 schema、有界结果、错误和 render intent。
-- TUI、sequence、BEL、auto-start、Windows 和 crash-restoration 行为不出现在公共 schema 中，并记录为推迟事项，而不是由 fixture 模拟。
-- 包 README 与 JSDoc 记录配置、所有权、失败、取消、上限、沙箱、模型可见影响和限制；实现同时更新 `docs/architecture.md` 与生成目录。
-- 根 `AGENTS.md` 中的仓库 CI 等价序列通过，包括 `test:coverage`、快照、文档、构建、hygiene 和 built-entry smoke。
+- 每文件覆盖率固定 owner 隔离、并发预留、生命周期清理、就绪层级、sanitizer carry state、UTF-8 上限、task 集成、schema 和 render intent。
+- Linux 进程 fixture 覆盖非 leader 与非主线程的 stdin 等待、不可读进程状态、受支持的 syscall 表、不支持的架构和误报拒绝；同一单元测试套件通过注入覆盖 macOS 检查器逻辑。
+- 真实 `node-pty` 测试在受支持宿主上覆盖 shell 状态、共享沙箱策略、环境清洗、信号、忽略 `SIGTERM` 的子进程，以及 dispose 返回后立即静默。
+- Loader 驱动的 `cordis.yml` 测试挂载真实三包组合；ACP 与 headless 快照通过 opt-in overlay 固定 6 个 schema、有界结果、错误渲染和 terminal/generic card。
+- 包契约、架构图、核心数据结构、生成目录和 website API 描述同一个已发布接口。
+- 仓库 CI 等价序列负责类型、lint、覆盖率、快照、文档、构建、hygiene、demo 和 built-entry 验证。
 
-## 风险
+## 后果
+
+**无需削弱一次性工具即可获得持久终端状态。**Shell 与 REPL 状态可以跨工具调用保留，而 `bash`、`read`、`write` 和 `edit` 继续拥有更窄的校验、审批与回放契约。
 
 **Linux Tier 1 之外的 idle 都是启发式结果。**输出静默无法区分 prompt、sleep 和网络 I/O。类型化结果保留不确定性，有界 timeout、task 等待与信号让模型仍能掌握控制权。
 
 **持久状态可能偏离模型认知。**模型可能忘记 cwd 或活跃 REPL。会话摘要和保留输出有助恢复，但任何 prompt 都无法让状态持久化变成确定行为。
 
+**daemonized 子进程可能离开捕获树。**在 teardown 前 reparent 的进程无法再从 `node-pty` 根进程发现。实现接受这个清理缺口，不冒险按 SID 向无关进程发送信号。
+
 **Shell 可以造成外部副作用。**会话沙箱和环境清洗降低本地暴露，但无法撤销 push、API 调用或消息发送。无法容忍这些副作用的部署必须省略 PTY 或增加网络策略。
 
 **进程丢失会销毁终端状态。**进程内会话无法跨 harness crash 或 restart 存活，原始 scrollback 也不持久化。重要工作必须提交到文件或其他持久系统。