feat(ui): add dedicated TUI package

Move the interactive pi-tui front door into @deepseek-ai/dsh-tui while keeping @deepseek-ai/dsh-stdio line-oriented for pipes. Select the terminal package in the demo app, preserve logger ownership, and cover the production Loader composition with a PTY smoke test.
This commit is contained in:
Turtle
2026-07-17 12:01:37 +08:00
parent 4139e093dd
commit 1119c537d0
49 changed files with 2808 additions and 93 deletions

View File

@@ -5,11 +5,11 @@ Pre-composed plugin bundles a thin leaf `cordis.yml` loads instead of assembling
| Package | npm name | Role |
|---|---|---|
| `agent-spine-demo/` | `@deepseek-ai/dsh-agent-spine-demo` | The executor-less/UI-less agent spine as one bundle plugin (`timer` + `llm` + sessions + system-prompt + tools + skills + agents + invariants + `tool-bash` + `tool-skill` + `agent-loop`) |
| `stdio-demo/` | `@deepseek-ai/dsh-stdio-demo` | Terminal stdio chat app: the spine + console logger + readline UI + a pre-created `main` agent, with a boot `bin` |
| `stdio-demo/` | `@deepseek-ai/dsh-stdio-demo` | Terminal chat app: the spine + JSONL persistence + TTY-selected `dsh-tui`/`dsh-stdio` front door + a pre-created `main` agent, with a boot `bin` |
| `acp-demo/` | `@deepseek-ai/dsh-acp-demo` | ACP server app: the spine + JSONL persistence + the [`acp`](../ui/acp/README.md) bridge (no stdout logger), with a boot `bin` |
| `jsonrpc-demo/` | `@deepseek-ai/dsh-jsonrpc-demo` | Bin-only runtime that boots an external `cordis.yml` for the stdio JSON-RPC SDK client |
`agent-spine-demo` is the shared bundle; `stdio-demo` and `acp-demo` compose it with opposite front-door clusters (console logger + readline UI vs the stdout-owning ACP bridge) and own their boot bins. `jsonrpc-demo` mounts no composition of its own — it boots whatever tree the deployment's `cordis.yml` names, and is what the Python SDK runtime launches.
`agent-spine-demo` is the shared bundle; `stdio-demo` and `acp-demo` compose it with terminal and ACP front-door clusters and own their boot bins. `jsonrpc-demo` mounts no composition of its own — it boots whatever tree the deployment's `cordis.yml` names, and is what the Python SDK runtime launches.
These are **not** product API. The spine pieces they bundle live in [`core/`](../core/README.md), the bridges/channels/boot-glue in [`ui/`](../ui/README.md), and the swappable backends (LLM adapter, bash executor) in their capability groups; a demo bundle just picks one concrete composition of them. Swap or fork one freely.

View File

@@ -33,7 +33,7 @@ The spine is everything COMMON to every front door. The swappable and front-door
- **the LLM adapter** — the bundle ships the abstract `llm` service; the leaf registers a concrete adapter on `ctx.llm` (`llm-deepseek`, `llm-pi-ai`, `llm-replay`).
- **the bash executor** — the bundle ships `tool-bash` (the consumer schema); the leaf provides `ctx.bash` (`bash-local` or a sandboxed impl).
- **non-local skill providers** — the bundle ships the skill registry, the local filesystem provider, and the `skill` tool; deployments can add other providers such as embedded or remote catalogs as siblings.
- **presentation + per-app infra** — the stdio UI / ACP bridge, a console logger, `hmr`. These form the coupled "front-door cluster" that the app packages ([`dsh-stdio-demo`](../../examples/stdio-demo/README.md), [`dsh-acp-demo`](../../examples/acp-demo/README.md)) bake in. `timer` is in the spine (common to both, stdout-silent); a console logger is NOT (it writes to stdout, which the ACP bridge reserves for JSON-RPC).
- **presentation + per-app infra** — the terminal (`dsh-tui` / `dsh-stdio`) or ACP front door and `hmr`. These form the coupled front-door cluster that the app packages ([`dsh-stdio-demo`](../../examples/stdio-demo/README.md), [`dsh-acp-demo`](../../examples/acp-demo/README.md)) bake in. `timer` is in the spine because it is common and stdout-silent; front doors own stdout and remain outside.
This is the [interface/implementation/consumer seam](../../../docs/rfc/implemented/architecture/2026-06-13-capability-seams.md) raised to the composition level: the bundle owns the shared spine, the leaf owns the backends, the app package owns the front door.

View File

@@ -1,8 +1,8 @@
# @deepseek-ai/dsh-stdio-demo
The **terminal stdio chat app**: a Cordis app plugin that composes the default agent spine ([`@deepseek-ai/dsh-agent-spine-demo`](../../examples/agent-spine-demo/README.md)) with the front-door cluster a terminal chat needs, and a `bin` that boots a leaf `cordis.yml`.
The **terminal chat app**: a Cordis app plugin that composes the default agent spine ([`@deepseek-ai/dsh-agent-spine-demo`](../../examples/agent-spine-demo/README.md)) with JSONL persistence, human interaction, a pre-created `main` agent, and a TTY-selected pi-tui/readline terminal front door. Its `bin` boots a leaf `cordis.yml`.
It is the readline counterpart to [`@deepseek-ai/dsh-acp-demo`](../acp-demo/README.md): both consume the same spine, but each bakes in the OPPOSITE front-door cluster.
It is the terminal counterpart to [`@deepseek-ai/dsh-acp-demo`](../acp-demo/README.md): both consume the same spine, while ACP reserves stdout for JSON-RPC and creates sessions from the client.
## What it bakes in
@@ -10,14 +10,15 @@ A terminal chat always wants the same cluster, so the package owns it rather tha
| Plugin | Why it is here |
|---|---|
| `@cordisjs/plugin-logger-console` | the console logger — stdout is just the terminal here, so logging to it is correct (the ACP app must NOT have this) |
| `@deepseek-ai/dsh-agent-spine-demo` | the spine, pre-creating a `main` agent from this app's `model` with `process.cwd()` as the fresh session cwd and carrying its `persona` |
| `@deepseek-ai/dsh-session-persistence-jsonl` | durable JSONL session log under `persistenceRoot` |
| `@deepseek-ai/dsh-user-interaction` | the human question/answer seam used by confirmation tools |
| `@deepseek-ai/dsh-tool-ask-user` | the model-facing `ask_user_question` tool |
| `@deepseek-ai/dsh-stdio` | the readline UI, bound to the `main` agent |
| `@cordisjs/plugin-logger-console` | readline diagnostics for non-TTY operation; omitted from the fullscreen TUI path |
| `@deepseek-ai/dsh-stdio` | the line-oriented terminal channel, bound to `main` for pipes and automation |
| `@deepseek-ai/dsh-tui` | the interactive pi-tui channel, bound to `main` for TTY pairs |
`@cordisjs/plugin-hmr` (the dev/demo edit-reload loop) is deliberately a **leaf** entry, NOT baked in here: it is a Loader-only, subprocess-only dev plugin — its constructor throws without `node --expose-internals` + a live `loader`, and the in-process test tier cannot even import it (so a package whose `apply` statically pulled it in could never carry the per-file coverage gate). Unlike the console logger, a stray `hmr` is not a stdout-purity footgun, so leaving it at the leaf costs no safety. The `demo:echo` / `demo:repl` leaves load it and pass `--expose-internals`.
`@cordisjs/plugin-hmr` (the dev/demo edit-reload loop) is deliberately a **leaf** entry, not baked in here: it is a Loader-only, subprocess-only dev plugin whose constructor needs `node --expose-internals` plus a live `loader`. The `demo:echo` / `demo:repl` leaves load it and pass `--expose-internals`.
The leaf `cordis.yml` supplies only the **swappable backends** — an LLM adapter (`llm-deepseek` for the real model, or the mock `mock-llm` for a demo) and a bash executor (`bash-local`) — `hmr`, plus this app's [`Config`](#config). The whole plugin tree a run loads is therefore: this app's cluster, the spine inside `agent-core`, `hmr`, and the two leaf backends.
@@ -33,7 +34,8 @@ The leaf `cordis.yml` supplies only the **swappable backends** — an LLM adapte
| `toolBash` | owner defaults | model-facing bash config routed through `dsh-agent-spine-demo`, including bash's producer-local `enableRunInBackground` |
| `toolTasks` | owner defaults | generic `task_output` wait bounds routed through `dsh-agent-spine-demo` |
| `persistenceRoot` | `./.sessions` | the JSONL backend's root directory |
| `welcome` | `ready.` | the stdin-chat banner |
| `welcome` | `ready.` | terminal banner / TUI subtitle |
| `ui` | owner defaults | app mode selection and nested `dsh-tui` presentation config |
| `resumeSessionId` | — | resume a persisted session id instead of starting fresh (sourced from an env var in the leaf) |
Fresh stdio sessions use the process launch directory as `session.header.cwd`, so project-scoped features such as skill discovery and default bash workdir follow the directory where `dsh-stdio-demo` was started. Resumed sessions keep the cwd stored in the persisted session header.
@@ -45,7 +47,7 @@ Fresh stdio sessions use the process launch directory as `session.header.cwd`, s
## Example leaf `cordis.yml`
```yaml
# A REPL agent demo: hmr + the DeepSeek adapter + local bash, then this app.
# A coding-agent demo: hmr + the DeepSeek adapter + local bash, then this app.
- id: hmr
name: '@cordisjs/plugin-hmr'
config:
@@ -64,6 +66,8 @@ Fresh stdio sessions use the process launch directory as `session.header.cwd`, s
config:
model: deepseek-v4-flash
persona: 'You are a coding assistant powered by the {{model}} model.'
ui:
mode: auto
```
Swap `llm-deepseek` for a `mock-llm` leaf plugin and you have the echo demo — "swap the backend, keep the app".
@@ -72,9 +76,9 @@ Swap `llm-deepseek` for a `mock-llm` leaf plugin and you have the echo demo —
### Composed terminal agent request
**What the model sees**: Through `dsh-agent-spine-demo`, the `main` agent receives the harness identity, configured persona, skill catalog, and visible tools; this app also composes the generated [`ask_user_question` schema](../../../docs/tool-catalog.md#deepseek-aidsh-tool-ask-user). Each readline submission becomes a user message.
**What the model sees**: Through `dsh-agent-spine-demo`, the `main` agent receives the harness identity, configured persona, skill catalog, and visible tools; this app also composes the generated [`ask_user_question` schema](../../../docs/tool-catalog.md#deepseek-aidsh-tool-ask-user). Each terminal submission becomes a user message; submissions made while the agent runs steer the active turn.
**Token effect**: Child prompt and schema costs repeat per request; user input and tool history grow until compaction. The welcome banner, logger output, and rendered transcript are terminal-only and add zero model tokens.
**Token effect**: Child prompt and schema costs repeat per request; user input and tool history grow until compaction. The TUI/readline banners and rendered transcripts are terminal-only and add zero model tokens.
### Human-answer result
@@ -84,6 +88,6 @@ Swap `llm-deepseek` for a `mock-llm` leaf plugin and you have the echo demo —
## Known Limitations and Deferred Work
- **One pre-created `main` agent drives the readline UI** — there is no multi-session or concurrent-agent surface in this app; a run is one conversation.
- **One pre-created `main` agent drives the terminal UI** — there is no multi-session or concurrent-agent surface in this app; a run is one conversation.
- **The front-door cluster is fixed in code** — the JSONL persistence backend and the ask-user tooling are baked; a different composition is a leaf-level sibling entry or another app package.
- **The question tool is not an approval answerer** — this app mounts `user-interaction` and `ask_user_question`, but not `ctx.approval`; a `tools/pre-execute` `ask` therefore fails closed unless the leaf composes an approval service and terminal answerer.

View File

@@ -1,6 +1,6 @@
{
"name": "@deepseek-ai/dsh-stdio-demo",
"description": "Terminal stdio chat app: the agent-spine-demo bundle + console logger + readline UI + a pre-created main agent, with a bin to boot a leaf cordis.yml",
"description": "Terminal chat app: agent spine + JSONL persistence + TTY pi-tui/readline front-door selection + pre-created main agent",
"version": "0.0.1",
"private": true,
"type": "module",
@@ -32,14 +32,15 @@
"peerDependencies": {
"@cordisjs/plugin-include": "^1.0.4",
"@cordisjs/plugin-loader": "^1.0.0-rc.5",
"@deepseek-ai/dsh-app-boot": "^0.0.1",
"@cordisjs/plugin-logger-console": "^1.0.0",
"@deepseek-ai/dsh-app-boot": "^0.0.1",
"@deepseek-ai/dsh-agent": "^0.0.1",
"@deepseek-ai/dsh-llm": "^0.0.1",
"@deepseek-ai/dsh-agent-spine-demo": "^0.0.1",
"@deepseek-ai/dsh-session": "^0.0.1",
"@deepseek-ai/dsh-session-persistence-jsonl": "^0.0.1",
"@deepseek-ai/dsh-stdio": "^0.0.1",
"@deepseek-ai/dsh-tui": "^0.0.1",
"@deepseek-ai/dsh-tool-ask-user": "^0.0.1",
"@deepseek-ai/dsh-tools": "^0.0.1",
"@deepseek-ai/dsh-user-interaction": "^0.0.1",
@@ -49,8 +50,8 @@
"devDependencies": {
"@cordisjs/plugin-include": "workspace:^",
"@cordisjs/plugin-loader": "workspace:^",
"@deepseek-ai/dsh-app-boot": "workspace:^",
"@cordisjs/plugin-logger-console": "workspace:^",
"@deepseek-ai/dsh-app-boot": "workspace:^",
"@deepseek-ai/dsh-agent": "workspace:^",
"@deepseek-ai/dsh-llm": "workspace:^",
"@deepseek-ai/dsh-agent-spine-demo": "workspace:^",
@@ -58,6 +59,7 @@
"@deepseek-ai/dsh-session": "workspace:^",
"@deepseek-ai/dsh-session-persistence-jsonl": "workspace:^",
"@deepseek-ai/dsh-stdio": "workspace:^",
"@deepseek-ai/dsh-tui": "workspace:^",
"@deepseek-ai/dsh-tool-ask-user": "workspace:^",
"@deepseek-ai/dsh-tools": "workspace:^",
"@deepseek-ai/dsh-user-interaction": "workspace:^",

View File

@@ -2,7 +2,7 @@
/**
* Boot a stdio app from a leaf `cordis.yml`; usage is `dsh-stdio-demo [config]`, defaulting to the
* cwd file. Shared `.env` loading, fail-loud Loader guards, and settled-tree boot live in
* dsh-app-boot. The echo and REPL demos invoke this bin with their own leaf configs.
* dsh-app-boot. The echo and coding-agent demos invoke this bin with their own leaf configs.
* @module @deepseek-ai/dsh-stdio-demo/bin
*/

View File

@@ -1,8 +1,9 @@
/**
* The stdio chat app: the default agent spine ({@link @deepseek-ai/dsh-agent-spine-demo}) plus the
* coupled front-door cluster a terminal chat needs — a console logger, the independently
* packaged readline UI, JSONL session persistence, the user-interaction seam with its
* `ask_user_question` tool, and a pre-created `main` agent the UI drives.
* coupled front-door cluster a terminal chat needs — the independently packaged
* pi-tui and readline front doors, JSONL session persistence, the user-interaction
* seam with its `ask_user_question` tool, and a pre-created `main` agent the UI
* drives. Interactive terminals use `dsh-tui`; pipes use `dsh-stdio` plus logging.
* Swappable adapters, executors, optional tools, and HMR stay in the leaf. This
* Loader plugin intentionally exposes named exports only; a default export
* would hide its `Config` schema (see docs/postmortem/0001).
@@ -20,9 +21,45 @@ import SessionPersistenceJsonl from '@deepseek-ai/dsh-session-persistence-jsonl'
import UserInteractionService from '@deepseek-ai/dsh-user-interaction'
import * as toolAskUser from '@deepseek-ai/dsh-tool-ask-user'
import * as uiStdio from '@deepseek-ai/dsh-stdio'
import * as uiTui from '@deepseek-ai/dsh-tui'
export const name = 'stdio-demo'
/** Terminal front door selected by the app bundle. */
export type TerminalMode = 'auto' | 'readline' | 'tui'
/** App-level terminal selection with nested TUI presentation settings. */
export interface UiConfig {
/** Select a concrete front door or infer it from the process streams. */
mode?: TerminalMode
/** Settings forwarded only when the pi-tui front door is selected. */
tui?: uiTui.TuiConfig
}
const terminalModeSchema = z.union(['auto', 'readline', 'tui'] as const).default('auto')
/** Schemastery schema for the app-level terminal selection. */
export const UiConfigSchema: z<UiConfig> = z.object({
mode: terminalModeSchema,
tui: uiTui.TuiConfigSchema,
})
/**
* Resolve the app's terminal front door.
*
* @param config - App-level terminal selection.
* @param isTTY - Whether both process streams are interactive TTYs.
* @returns The concrete UI package to mount.
*/
export function resolveTerminalMode(config: UiConfig | undefined, isTTY: boolean): Exclude<TerminalMode, 'auto'> {
const mode = config?.mode ?? 'auto'
if (mode === 'auto') return isTTY ? 'tui' : 'readline'
if (mode === 'tui' && !isTTY) {
throw new Error('stdio-demo: TUI mode requires both stdin and stdout to be TTYs; use mode "readline" for pipes')
}
return mode
}
/**
* App config: the swappable per-demo values, each routed to where the app wires
* it. `model`/`resumeSessionId` configure the pre-created `main` agent (through
@@ -31,7 +68,7 @@ export const name = 'stdio-demo'
* is the explicit model-facing tool order (forwarded to the system-prompt plugin);
* fresh sessions use `process.cwd()` as their workspace cwd; resumed sessions
* keep their persisted cwd. `persistenceRoot` is the JSONL backend's directory;
* `welcome` is the UI banner.
* `welcome` is the UI banner and `ui` configures terminal mode/presentation.
*/
export interface Config {
/** Model name for the `main` agent (must have a registered adapter). */
@@ -44,8 +81,10 @@ export interface Config {
tools?: ToolsConfig
/** Directory the JSONL session backend writes under. Defaults to `./.sessions`. */
persistenceRoot?: string
/** stdin-chat banner printed once on start. Defaults to `'ready.'`. */
/** Terminal banner printed once on start. Defaults to `'ready.'`. */
welcome?: string
/** Terminal front-door selection and pi-tui presentation settings. */
ui?: UiConfig
/** Skill registry, local-provider, and model-facing consumer config forwarded to agent-spine-demo. */
skills?: agentCore.SkillConfig
/** Model-facing bash tool config forwarded through agent-core. */
@@ -72,6 +111,7 @@ export const Config: z<Config> = z.object({
// apply() fallbacks through named constants while retaining both boundaries.
persistenceRoot: z.string().default('./.sessions'),
welcome: z.string().default('ready.'),
ui: UiConfigSchema,
skills: agentCore.SkillConfigSchema,
toolBash: agentCore.ToolBashConfigSchema,
toolTasks: agentCore.ToolTasksConfigSchema,
@@ -79,14 +119,18 @@ export const Config: z<Config> = z.object({
})
/**
* Compose the spine with the stdio front door. The console logger comes first
* (infra), then the agent-spine-demo bundle pre-creating the `main` agent from this
* app's `model`/`resumeSessionId` with the deployment `persona`, then the JSONL
* backend, then the readline UI bound to `main`. The `hmr` dev-reload plugin is
* a leaf concern (see the module doc), so it is not mounted here.
* Compose the spine with one terminal front door. Interactive TTY pairs mount
* `dsh-tui` without a console exporter; pipes mount the readline `dsh-stdio`
* channel with the console logger. The `hmr` dev-reload plugin remains a leaf
* concern.
*
* @param ctx - Context receiving the app's child plugins.
* @param config - App configuration routed to the spine and front door.
* @param isTTY - Whether both terminal streams are interactive TTYs.
*/
export function apply(ctx: Context, config: Config): void {
ctx.plugin(ConsoleExporter)
export function composeTerminalApp(ctx: Context, config: Config, isTTY: boolean): void {
const mode = resolveTerminalMode(config.ui, isTTY)
if (mode === 'readline') ctx.plugin(ConsoleExporter)
ctx.plugin(agentCore, {
...config.persona !== undefined ? { persona: config.persona } : {},
...config.toolOrder !== undefined ? { toolOrder: config.toolOrder } : {},
@@ -104,5 +148,21 @@ export function apply(ctx: Context, config: Config): void {
ctx.plugin(SessionPersistenceJsonl, { root: config.persistenceRoot ?? './.sessions' })
ctx.plugin(UserInteractionService)
ctx.plugin(toolAskUser)
ctx.plugin(uiStdio, { welcome: config.welcome ?? 'ready.', agent: 'main' })
if (mode === 'tui') {
ctx.plugin(uiTui, {
...config.ui?.tui,
welcome: config.welcome ?? 'ready.',
agent: 'main',
})
} else {
ctx.plugin(uiStdio, { welcome: config.welcome ?? 'ready.', agent: 'main' })
}
}
/** Compose the configured terminal front door with the agent app. */
/* v8 ignore start -- production stream capability wiring; composeTerminalApp is unit-covered,
and the coding-agent PTY smoke covers the interactive process path */
export function apply(ctx: Context, config: Config): void {
composeTerminalApp(ctx, config, process.stdin.isTTY && process.stdout.isTTY)
}
/* v8 ignore stop */

View File

@@ -10,8 +10,8 @@ import { TOOL_ORDER_REST } from '@deepseek-ai/dsh-system-prompt'
import * as stdioAgent from '../src/index.ts'
/**
* Unit coverage for app composition and config forwarding: console logger, pre-created main agent,
* agent-core spine, JSONL backend, and readline UI. HMR is a Loader-only leaf concern covered by the
* Unit coverage for app composition and config forwarding: pre-created main agent, agent-core spine,
* JSONL backend, and adaptive terminal UI. HMR is a Loader-only leaf concern covered by the
* keyless echo smoke; this tier pins the export shape because an inject-less app could otherwise
* survive namespace collapse while silently losing its schema.
*/
@@ -65,6 +65,47 @@ async function withIsolatedSkillHomes<T>(run: () => Promise<T>): Promise<T> {
}
describe('dsh-stdio-demo app', () => {
it('selects readline for pipes and dsh-tui for interactive terminal pairs', () => {
expect(stdioAgent.resolveTerminalMode(undefined, false)).toBe('readline')
expect(stdioAgent.resolveTerminalMode(undefined, true)).toBe('tui')
expect(stdioAgent.resolveTerminalMode({ mode: 'readline' }, true)).toBe('readline')
expect(stdioAgent.resolveTerminalMode({ mode: 'tui' }, true)).toBe('tui')
expect(() => stdioAgent.resolveTerminalMode({ mode: 'tui' }, false)).toThrow('requires both stdin and stdout')
})
it('composes only the selected terminal package and keeps TUI settings in dsh-tui', () => {
const calls: Array<{ name: string; config: unknown }> = []
const ctx = {
plugin(plugin: { name?: string }, config?: unknown) {
calls.push({ name: plugin.name ?? '', config })
},
} as unknown as Context
stdioAgent.composeTerminalApp(ctx, {
model: 'mock',
welcome: 'TUI ready',
ui: { mode: 'tui', tui: { color: false, maxToolOutputLines: 3 } },
}, true)
expect(calls.map(call => call.name)).toContain('ui-tui')
expect(calls.map(call => call.name)).not.toContain('ui-stdio')
expect(calls.map(call => call.name)).not.toContain('ConsoleExporter')
expect(calls.find(call => call.name === 'ui-tui')?.config).toMatchObject({
agent: 'main', welcome: 'TUI ready', color: false, maxToolOutputLines: 3,
})
calls.length = 0
stdioAgent.composeTerminalApp(ctx, { model: 'mock', ui: { mode: 'tui' } }, true)
expect(calls.find(call => call.name === 'ui-tui')?.config).toMatchObject({
agent: 'main', welcome: 'ready.',
})
calls.length = 0
stdioAgent.composeTerminalApp(ctx, { model: 'mock', ui: { mode: 'readline' } }, false)
expect(calls.map(call => call.name)).toContain('ui-stdio')
expect(calls.map(call => call.name)).toContain('ConsoleExporter')
expect(calls.map(call => call.name)).not.toContain('ui-tui')
})
it('composes the spine + front-door cluster and pre-creates the main agent', async () => {
const ctx = await mount({ model: 'mock', persona: 'hi', persistenceRoot: '/tmp/dsh-stdio-demo-spec', skills: await isolatedSkillsConfig() })
// The spine services (brought up by the agent-core bundle) are all present.

View File

@@ -38,6 +38,9 @@
{
"path": "../../ui/stdio"
},
{
"path": "../../ui/tui"
},
{
"path": "../../ui/tool-ask-user"
},

View File

@@ -6,4 +6,4 @@ The model-facing todo tool. A single **product** package — there is no interfa
|---|---|---|
| `tool-todo/` | Model-facing `todo_write` tool; writes the whole list to the session log (`todo/write`) | (registers on `ctx.tools`) |
The list lives on the event-sourced session log (`SessionEventMap['todo/write']`, owned by [`dsh-session`](../core/session)); this package is the thin consumer that appends the snapshot. UIs render off `session/event`: the [stdio app's readline UI](../examples/stdio-demo) prints the list, the [ACP bridge](../ui/acp) maps it to a `plan` sessionUpdate.
The list lives on the event-sourced session log (`SessionEventMap['todo/write']`, owned by [`dsh-session`](../core/session)); this package is the thin consumer that appends the snapshot. UIs render off `session/event`: the [terminal app](../examples/stdio-demo) shows a persistent TUI plan or readline checklist, while the [ACP bridge](../ui/acp) maps it to a `plan` sessionUpdate.

View File

@@ -18,7 +18,7 @@ Beyond the schema's type/required/enum checks, `execute` rejects an empty or dup
## Rendering
The tool writes only the session event; it does not render. UIs subscribe to `session/event` and render the `todo/write` data themselves: the [stdio app's readline UI](../../examples/stdio-demo) prints a glyphed checklist, and the [ACP bridge](../../ui/acp) maps the list to a `plan` sessionUpdate (synthesizing the `priority` ACP requires).
The tool writes only the session event; it does not render. UIs subscribe to `session/event` and render the `todo/write` data themselves: the [terminal app](../../examples/stdio-demo) shows a persistent TUI plan or readline checklist, and the [ACP bridge](../../ui/acp) maps the list to a `plan` sessionUpdate (synthesizing the `priority` ACP requires).
## Export shape

View File

@@ -9,12 +9,13 @@ Integrations that expose the agent to an external editor or client. These are **
| `permission/` | User-facing permission presets (`workspace-write`/`danger-full-access`): one product-level select bundling the sandbox-mode and approval-policy knobs, written through to their session events | `ctx.permission` |
| `user-interaction/` | Abstract human question/answer seam used by UI-backed confirmation tools | `ctx.userInteraction` |
| `tool-ask-user/` | Model-facing `ask_user_question` tool over `ctx.userInteraction` | (registers on `ctx.tools`) |
| `stdio/` | Terminal readline channel over `ctx.agents`, `session/event`, and `ctx.userInteraction`; agent lifecycle stays with app/developer code | (drives `ctx.agents`) |
| `stdio/` | Line-oriented terminal channel for pipes and automation; drives `ctx.agents`, renders `session/event`, and answers `ctx.userInteraction` | (drives `ctx.agents`) |
| `tui/` | Interactive pi-tui terminal channel for TTY sessions; renders `session/event`, tool presentation intents, and answers `ctx.userInteraction` | (drives `ctx.agents`) |
| `jsonrpc/` | Stdio JSON-RPC server for out-of-process SDK clients | (drives `ctx.agents`) |
| `app-boot/` | Shared boot glue for the app bins: `.env` loading, fail-loud Loader guards, snapshot-aware config resolution, the settle-the-tree boot sequence | (library for the bins) |
A UI integration is a client-driver plugin, not a loop change and not a capability seam: it consumes the existing `agent/*` event taxonomy and the `dsh-agent` factory. The `jsonrpc` plugin is the SDK-client sibling of the `acp` bridge (a JSON-RPC server over `ctx.agents` for out-of-process SDK clients rather than editors). The [`stdio`](stdio/README.md) plugin is the unstructured readline analogue of the `acp` bridge; app bundles and SDK projects compose it explicitly with the services and tools their product profile selects.
A UI integration is a client-driver plugin, not a loop change and not a capability seam: it consumes the existing `agent/*` event taxonomy and the `dsh-agent` factory. The `jsonrpc` plugin is the SDK-client sibling of the `acp` bridge (a JSON-RPC server over `ctx.agents` for out-of-process SDK clients rather than editors). The [`stdio`](stdio/README.md) and [`tui`](tui/README.md) plugins are the two terminal front doors: one is line-oriented for pipes, the other is interactive for TTYs. App bundles and SDK projects compose the appropriate channel explicitly with the services and tools their product profile selects.
`user-approval`, `user-interaction`, and `tool-ask-user` live here because asking a human is a UI-backed product affordance, not part of the providerless core spine. `user-approval` owns the one-shot `ctx.approval` decision mechanism and its policy tier; answerers remain with their UI channel owners. `user-interaction` remains provider-neutral (`ctx.userInteraction`), while `tool-ask-user` is its model-facing consumer and the app/bridge packages provide concrete providers.
The runnable app bundles that bake these bridges into boot bins — the stdio chat app, the ACP server app, and the JSON-RPC SDK-runtime bin — live in [`examples/`](../examples/README.md) (`stdio-demo`, `acp-demo`, `jsonrpc-demo`), each composed over the [`agent-spine-demo`](../examples/agent-spine-demo/README.md) bundle. `ui/` keeps the reusable bridge/channel plugins and the `app-boot` glue; each front door owns its stdout policy, and a leaf `cordis.yml` supplies backends and optional tools.
The runnable app bundles that bake these bridges into boot bins — the terminal chat app, the ACP server app, and the JSON-RPC SDK-runtime bin — live in [`examples/`](../examples/README.md) (`stdio-demo`, `acp-demo`, `jsonrpc-demo`), each composed over the [`agent-spine-demo`](../examples/agent-spine-demo/README.md) bundle. `ui/` keeps the reusable bridge/channel plugins and the `app-boot` glue; each front door owns its stdout policy, and a leaf `cordis.yml` supplies backends and optional tools.

View File

@@ -2,7 +2,7 @@
Agent Client Protocol bridge over JSON-RPC stdio. Editors can create or resume agents, stream their events, answer questions and approvals, and render tool calls. One connection supports multiple isolated sessions; Zed is the primary compatibility target.
It is a **client-driver / UI plugin**, the structured analogue of the readline `stdio-chat` plugin — NOT a loop change and NOT a [capability seam](../../../docs/rfc/implemented/architecture/2026-06-13-capability-seams.md). It consumes the existing `agent/*` event taxonomy, the `dsh-agent` create/resume factory, and `dsh-session-persistence`.
It is a **client-driver / UI plugin**, the structured analogue of the terminal `dsh-tui`/`dsh-stdio` channels — NOT a loop change and NOT a [capability seam](../../../docs/rfc/implemented/architecture/2026-06-13-capability-seams.md). It consumes the existing `agent/*` event taxonomy, the `dsh-agent` create/resume factory, and `dsh-session-persistence`.
## Service / plugin

56
packages/ui/tui/README.md Normal file
View File

@@ -0,0 +1,56 @@
# @deepseek-ai/dsh-tui
The interactive terminal front door for DeepSeek Harness agents, built on [`@earendil-works/pi-tui`](https://www.npmjs.com/package/@earendil-works/pi-tui). It requires stdin and stdout TTYs; scripts and Loader pipes should compose [`@deepseek-ai/dsh-stdio`](../stdio/README.md) instead.
This package owns interactive terminal presentation and input only. It injects `agents`, `tools`, and `userInteraction`, then drives an agent created or resumed by app or developer code. Agent lifecycle, persistence, and the model-facing [`ask_user_question`](../tool-ask-user/README.md) tool remain separate composition entries.
The TUI rebuilds resumed history from the active session surface, renders Markdown responses and reasoning, applies each tool's `presentCall` / `presentResult` intent to terminal, diff, or generic cards, keeps the latest `todo/write` plan above the editor, and presents `ctx.userInteraction` questions as keyboard-driven overlays. Surface replacement events rebuild the transcript so compacted history does not reappear.
While the agent is running, editor submissions call `agent.steer()`; otherwise they call `agent.send()`. Ctrl+C or Escape cancels a running turn. Ctrl+O expands tool cards, Ctrl+R toggles reasoning, Ctrl+L redraws, and Ctrl+D exits while idle. `/help`, `/clear`, `/cancel`, `/reasoning`, `/tools`, `/redraw`, and `/exit` provide the same actions without key chords.
## Config
| Key | Default | Meaning |
|---|---|---|
| `welcome` | `ready.` | Header subtitle |
| `agent` | `main` | Agent id driven by the terminal |
| `showReasoning` | `true` | Render reasoning blocks |
| `maxToolOutputLines` | `12` | Collapsed tool-card output limit |
| `maxQuestionOptions` | `8` | Visible options in a question overlay |
| `questionDialogWidth` | `72` | Question-overlay width in columns |
| `questionDialogMaxHeight` | `20` | Question-overlay maximum rows |
| `showHardwareCursor` | `false` | Show the hardware cursor at pi-tui's IME marker |
| `color` | `true` | Apply the built-in ANSI palette |
| `title` | `DeepSeek Harness` | Terminal window title |
```yaml
- id: terminal
name: '@deepseek-ai/dsh-tui'
config:
welcome: 'Coding agent ready.'
agent: main
showReasoning: true
maxToolOutputLines: 12
```
Startup fails before mounting when either process stream is not a TTY. Disposal stops loaders, rejects pending questions, drains terminal input, restores terminal state, unregisters event listeners and the user-interaction provider, and never exits a replacement process during HMR.
## Model Experience
### Interactive prompt input
**What the model sees**: Each non-empty editor submission becomes one text block, sent with `agent.send()` while the target agent is idle and `agent.steer()` while it is running. Slash commands and keybindings are TUI-only.
**Token effect**: Submitted text is retained under the agent loop's normal session-history and compaction rules. Headers, cards, Markdown rendering, status lines, plans, and help text add no tokens.
### Interactive user-question answers
**What the model sees**: When a consumer calls `ctx.userInteraction.ask()`, this provider presents each question in order and returns selected option labels or `custom` text. Abort, cancellation, or UI disposal becomes `Error: ask_user_question was interrupted before the user answered` through `dsh-tool-ask-user`.
**Token effect**: Waiting and terminal overlays add no tokens; the resolved answer or error is model-visible only through the calling tool or plugin's result.
## Known Limitations and Deferred Work
- **One configured agent owns the transcript and editor** — questions from other agents can still use the shared overlay provider, but session rendering and prompt input remain bound to `agent`.
- **Tool cards are text terminal presentations** — terminal, diff, and generic cards use tool-owned titles/content, but session content currently has no image block for inline image rendering.
- **Non-TTY operation is intentionally unsupported** — app bundles that need automation must select `dsh-stdio` before mounting this plugin rather than expecting an internal fallback.

View File

@@ -0,0 +1,45 @@
{
"name": "@deepseek-ai/dsh-tui",
"description": "Interactive pi-tui terminal front door for DeepSeek Harness agents",
"version": "0.0.1",
"private": true,
"type": "module",
"main": "lib/index.js",
"types": "lib/types/index.d.ts",
"exports": {
".": {
"types": "./lib/types/index.d.ts",
"default": "./lib/index.js"
},
"./src/*": "./src/*",
"./package.json": "./package.json"
},
"files": [
"lib/index.js",
"lib/types/**/*.d.ts",
"lib/types/**/*.d.ts.map",
"src"
],
"license": "BSD-3-Clause",
"peerDependencies": {
"@deepseek-ai/dsh-agent": "^0.0.1",
"@deepseek-ai/dsh-llm": "^0.0.1",
"@deepseek-ai/dsh-session": "^0.0.1",
"@deepseek-ai/dsh-tools": "^0.0.1",
"@deepseek-ai/dsh-user-interaction": "^0.0.1",
"cordis": "^4.0.0-rc.6"
},
"dependencies": {
"@earendil-works/pi-tui": "0.80.7",
"schemastery": "^3.18.0"
},
"devDependencies": {
"@cordisjs/plugin-loader": "workspace:^",
"@deepseek-ai/dsh-agent": "workspace:^",
"@deepseek-ai/dsh-llm": "workspace:^",
"@deepseek-ai/dsh-session": "workspace:^",
"@deepseek-ai/dsh-tools": "workspace:^",
"@deepseek-ai/dsh-user-interaction": "workspace:^",
"cordis": "^4.0.0-rc.6"
}
}

1284
packages/ui/tui/src/index.ts Normal file

File diff suppressed because it is too large Load Diff

View File

@@ -0,0 +1,19 @@
import { describe, expect, it } from 'vitest'
import Loader from '@cordisjs/plugin-loader'
import * as tui from '../src/index.ts'
/** Real Loader export-path guard for the namespace TUI plugin. */
describe('dsh-tui plugin export shape', () => {
it('preserves name, inject, Config, and apply through Loader unwrapping', () => {
expect('default' in tui).toBe(false)
expect(typeof tui.apply).toBe('function')
const loader = Object.create(Loader.prototype) as Loader
const unwrapped = loader.unwrapExports(tui) as Record<string, unknown>
expect(unwrapped).toBe(tui)
expect(unwrapped.name).toBe('ui-tui')
expect(unwrapped.inject).toEqual(['agents', 'userInteraction', 'tools'])
expect(unwrapped.Config).toBeDefined()
expect(typeof unwrapped.apply).toBe('function')
})
})

View File

@@ -0,0 +1,952 @@
import { describe, expect, it, vi } from 'vitest'
import { Context } from 'cordis'
import type { Terminal } from '@earendil-works/pi-tui'
import AgentRegistry, { AgentId, type Agent, type AgentStatus } from '@deepseek-ai/dsh-agent'
import type { ContentBlock } from '@deepseek-ai/dsh-llm'
import SessionStore, { SessionId, type Session } from '@deepseek-ai/dsh-session'
import type { ToolDefinition } from '@deepseek-ai/dsh-tools'
import UserInteractionService from '@deepseek-ai/dsh-user-interaction'
import {
createTuiChat,
mountTui,
resolveTuiConfig,
type Config,
type TuiRuntime,
} from '../src/index.ts'
class FakeTerminal implements Terminal {
columns = 88
rows = 32
kittyProtocolActive = false
output = ''
title = ''
progress: boolean[] = []
started = 0
stopped = 0
drainInput = vi.fn(() => Promise.resolve())
private onInput: (data: string) => void = () => {}
private onResize: () => void = () => {}
start(onInput: (data: string) => void, onResize: () => void): void {
this.started += 1
this.onInput = onInput
this.onResize = onResize
}
stop(): void {
this.stopped += 1
}
write(data: string): void {
this.output += data
}
moveBy(lines: number): void {
this.output += `[move:${lines}]`
}
hideCursor(): void {
this.output += '[hide]'
}
showCursor(): void {
this.output += '[show]'
}
clearLine(): void {
this.output += '[clear-line]'
}
clearFromCursor(): void {
this.output += '[clear-rest]'
}
clearScreen(): void {
this.output += '[clear-screen]'
}
setTitle(title: string): void {
this.title = title
}
setProgress(active: boolean): void {
this.progress.push(active)
}
send(data: string): void {
this.onInput(data)
}
resize(columns: number, rows = this.rows): void {
this.columns = columns
this.rows = rows
this.onResize()
}
}
interface FakeAgent extends Agent {
status: AgentStatus
sent: ContentBlock[][]
steered: ContentBlock[][]
cancelled: string[]
}
async function tick(): Promise<void> {
await new Promise(resolve => setTimeout(resolve, 25))
}
async function setup(options: {
status?: AgentStatus
config?: Config
tools?: Record<string, ToolDefinition>
beforeMount?: (session: Session) => void
cwd?: string | null
} = {}) {
const ctx = new Context()
await ctx.plugin(SessionStore)
await ctx.plugin(AgentRegistry)
await ctx.plugin(UserInteractionService)
const tools = options.tools ?? {}
ctx.provide('tools', {
get(name: string) {
return tools[name]
},
} as never)
const session = ctx.sessions.create(
SessionId('main-session'),
options.cwd === null ? undefined : { meta: { cwd: options.cwd ?? process.cwd() } },
)
options.beforeMount?.(session)
const sent: ContentBlock[][] = []
const steered: ContentBlock[][] = []
const cancelled: string[] = []
const agent: FakeAgent = {
id: AgentId('main'),
options: { model: 'deepseek-v4-flash' },
session,
status: options.status ?? 'idle',
ctx,
sent,
steered,
cancelled,
send(content) {
sent.push(content)
},
steer(content) {
steered.push(content)
},
inject() {},
cancel(reason) {
cancelled.push(reason ?? '')
},
whenIdle() {
return Promise.resolve()
},
}
ctx.agents.register(agent)
const terminal = new FakeTerminal()
const exit = vi.fn()
const controller = createTuiChat(ctx, Object.assign({
welcome: 'Coding agent ready.',
agent: 'main',
color: false,
}, options.config), { terminal, exit })
await tick()
return { ctx, session, agent, terminal, exit, controller }
}
async function dispose(setupResult: Awaited<ReturnType<typeof setup>>): Promise<void> {
await setupResult.controller.dispose()
await setupResult.ctx.fiber.dispose()
}
function appendUser(session: Session, text: string): void {
session.append('user/message', {
content: [{ type: 'text', text }],
source: { kind: 'user' },
}, { surfaceOp: 'append' })
}
function appendAssistant(session: Session, content: ContentBlock[], usage?: { inputTokens: number; outputTokens: number }): void {
session.append('assistant/message', {
turn: 1,
step: 0,
content,
...usage === undefined ? {} : { usage },
}, { surfaceOp: 'append' })
}
describe('TUI config', () => {
it('defaults every direct-call TUI option', () => {
expect(resolveTuiConfig(undefined)).toEqual({
showReasoning: true,
maxToolOutputLines: 12,
maxQuestionOptions: 8,
questionDialogWidth: 72,
questionDialogMaxHeight: 20,
showHardwareCursor: false,
color: true,
title: 'DeepSeek Harness',
})
expect(resolveTuiConfig({
showReasoning: false,
maxToolOutputLines: 2,
maxQuestionOptions: 3,
questionDialogWidth: 60,
questionDialogMaxHeight: 14,
showHardwareCursor: true,
color: false,
title: 'DSH',
})).toEqual({
showReasoning: false,
maxToolOutputLines: 2,
maxQuestionOptions: 3,
questionDialogWidth: 60,
questionDialogMaxHeight: 14,
showHardwareCursor: true,
color: false,
title: 'DSH',
})
})
})
describe('pi-tui chat lifecycle and transcript', () => {
it('renders its header, footer, replay, streaming answer, todos, and status', async () => {
const result = await setup({
beforeMount(session) {
appendUser(session, 'restored prompt')
appendAssistant(session, [
{ type: 'reasoning', text: 'restored thought' },
{ type: 'text', text: '**restored answer**' },
], { inputTokens: 1_250, outputTokens: 42 })
session.append('todo/write', {
todos: [
{ content: 'read code', status: 'completed' },
{ content: 'write tests', status: 'in_progress' },
{ content: 'ship', status: 'pending' },
],
})
},
})
expect(result.terminal.started).toBe(1)
expect(result.terminal.title).toBe('DeepSeek Harness')
expect(result.terminal.output).toContain('DEEPSEEK')
expect(result.terminal.output).toContain('Coding agent ready.')
expect(result.terminal.output).toContain('restored prompt')
expect(result.terminal.output).toContain('restored thought')
expect(result.terminal.output).toContain('restored answer')
expect(result.terminal.output).toContain('write tests')
expect(result.terminal.output).toContain('↑1.3k ↓42')
result.agent.status = 'running'
result.ctx.emit('agent/status', result.agent, 'running')
result.session.append('user/message', { content: [{ type: 'text', text: ' ' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
result.session.append('steering/message', { turn: 2, content: [{ type: 'text', text: 'steering note' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
result.session.append('steering/message', { turn: 2, content: [{ type: 'text', text: '' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
result.session.append('context/message', { content: [{ type: 'text', text: 'user context' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
result.session.append('context/message', { content: [{ type: 'text', text: '' }], source: { kind: 'user' } }, { surfaceOp: 'append' })
result.session.append('prompt/blocked', { content: [{ type: 'text', text: 'blocked' }], source: { kind: 'user' }, reason: 'test policy' })
appendAssistant(result.session, [])
result.session.append('turn/end', { turn: 9, reason: { kind: 'aborted' } })
result.session.append('turn/end', { turn: 10, reason: { kind: 'completed' } })
result.session.append('step/start', { turn: 11, step: 0 })
result.session.append('assistant/chunk', {
turn: 2,
step: 0,
chunk: { type: 'block-start', index: 0, blockType: 'reasoning' },
})
result.session.append('assistant/chunk', {
turn: 2,
step: 0,
chunk: { type: 'reasoning-delta', index: 0, text: 'live thought' },
})
result.session.append('assistant/chunk', {
turn: 2,
step: 0,
chunk: { type: 'reasoning-delta', index: 9, text: 'unannounced thought' },
})
result.session.append('assistant/chunk', {
turn: 2,
step: 0,
chunk: { type: 'block-end', index: 0, block: { type: 'reasoning', text: 'live thought complete' } },
})
result.session.append('assistant/chunk', {
turn: 2,
step: 0,
chunk: { type: 'block-start', index: 1, blockType: 'text' },
})
result.session.append('assistant/chunk', {
turn: 2,
step: 0,
chunk: { type: 'text-delta', index: 1, text: 'live answer' },
})
result.session.append('assistant/chunk', {
turn: 2,
step: 0,
chunk: { type: 'block-end', index: 1, block: { type: 'text', text: 'live answer done' } },
})
result.session.append('assistant/chunk', {
turn: 2,
step: 0,
chunk: { type: 'block-start', index: 2, blockType: 'tool-call' },
})
result.session.append('assistant/chunk', {
turn: 2,
step: 0,
chunk: { type: 'block-end', index: 2, block: { type: 'tool-call', id: 'stream-tool' as never, name: 'tool', arguments: '{}' } },
})
result.session.append('assistant/chunk', {
turn: 2,
step: 0,
chunk: { type: 'tool-call-delta', index: 2, id: 'stream-tool' as never, argumentsDelta: '{}' },
})
result.session.append('assistant/chunk', {
turn: 2,
step: 0,
chunk: { type: 'usage', usage: { inputTokens: 1, outputTokens: 2 } },
})
await tick()
expect(result.terminal.output).toContain('live thought')
result.terminal.send('\x12')
await tick()
appendAssistant(result.session, [{ type: 'text', text: 'final live answer' }], { inputTokens: 500, outputTokens: 8 })
await tick()
expect(result.terminal.output).toContain('Working')
expect(result.terminal.output).toContain('Steering')
expect(result.terminal.output).toContain('user context')
expect(result.terminal.output).toContain('Prompt blocked')
expect(result.terminal.output).toContain('Turn cancelled')
expect(result.terminal.output).toContain('final live answer')
expect(result.terminal.output).toContain('↑1.8k ↓50')
expect(result.terminal.progress).toContain(true)
result.session.append('assistant/chunk', {
turn: 3,
step: 0,
chunk: { type: 'text-delta', index: 0, text: 'cleared stream' },
})
result.terminal.send('/clear')
result.terminal.send('\r')
appendAssistant(result.session, [{ type: 'text', text: 'answer after clear' }])
await tick()
expect(result.terminal.output).toContain('answer after clear')
result.agent.status = 'idle'
result.ctx.emit('agent/status', result.agent, 'idle')
await tick()
expect(result.terminal.progress.at(-1)).toBe(false)
await dispose(result)
expect(result.terminal.stopped).toBe(1)
expect(result.terminal.drainInput).toHaveBeenCalledWith(100, 20)
})
it('renders the ANSI palette and every markdown/content style', async () => {
const result = await setup({
config: { color: true },
beforeMount(session) {
session.append('user/message', {
content: [
{ type: 'text', text: '# Heading\n\n[link](https://example.com) `code`\n\n```ts\nconst x = 1\n```\n\n> quote\n\n---\n\n- item\n\n**bold** *italic* ~~strike~~' },
{ type: 'tool-call', id: 'nested' as never, name: 'nested_tool', arguments: '{}' },
{ type: 'tool-result', toolCallId: 'nested' as never, content: [{ type: 'reasoning', text: 'nested result' }] },
{ type: 'future-block' } as never,
{} as never,
],
source: { kind: 'user' },
}, { surfaceOp: 'append' })
appendAssistant(session, [
{ type: 'reasoning', text: 'styled reasoning' },
{ type: 'text', text: 'styled answer' },
], { inputTokens: 2_000_000, outputTokens: 1_500_000 })
session.append('todo/write', { todos: [
{ content: 'done', status: 'completed' },
{ content: 'active', status: 'in_progress' },
{ content: 'later', status: 'pending' },
] })
},
})
result.terminal.send('/')
await tick()
result.terminal.send('zz')
await tick()
result.terminal.send('\x0c')
await tick()
expect(result.terminal.output).toContain('\x1b[')
expect(result.terminal.output).toContain('Heading')
expect(result.terminal.output).toContain('nested_tool({})')
expect(result.terminal.output).toContain('nested result')
expect(result.terminal.output).toContain('[future-block]')
expect(result.terminal.output).toContain('[content]')
expect(result.terminal.output).toContain('↑2.0m ↓1.5m')
await dispose(result)
})
it('suppresses stale replay chunks and does not duplicate editor history on rebuild', async () => {
const result = await setup({
beforeMount(session) {
appendUser(session, 'first prompt')
appendUser(session, 'second prompt')
session.append('assistant/chunk', {
turn: 2,
step: 0,
chunk: { type: 'text-delta', index: 0, text: 'stale partial response' },
})
},
})
expect(result.terminal.output).not.toContain('stale partial response')
result.terminal.send('/reasoning')
result.terminal.send('\r')
result.terminal.send('\x1b[A')
result.terminal.send('\x1b[A')
result.terminal.send('\x1b[A')
result.terminal.send('\r')
expect(result.agent.sent).toEqual([[{ type: 'text', text: 'first prompt' }]])
await dispose(result)
})
it('formats large token totals and cwd variants', async () => {
const homeResult = await setup({
cwd: process.env.HOME ?? process.cwd(),
beforeMount(session) {
appendAssistant(session, [{ type: 'text', text: 'home' }], { inputTokens: 25_000, outputTokens: 10_000 })
},
})
expect(homeResult.terminal.output).toContain('~ ↑25k ↓10k')
await dispose(homeResult)
const unsetResult = await setup({ cwd: null })
expect(unsetResult.terminal.output).toContain('cwd unset')
await dispose(unsetResult)
const outsideResult = await setup({ cwd: '/opt' })
expect(outsideResult.terminal.output).toContain('/opt')
await dispose(outsideResult)
})
it('sends, steers, handles commands, global keys, and disposed-agent input', async () => {
const result = await setup()
result.terminal.send('do the work')
result.terminal.send('\r')
expect(result.agent.sent).toEqual([[{ type: 'text', text: 'do the work' }]])
result.terminal.send(' ')
result.terminal.send('\r')
result.agent.status = 'running'
result.terminal.send('steer it')
result.terminal.send('\r')
expect(result.agent.steered).toEqual([[{ type: 'text', text: 'steer it' }]])
result.terminal.send('\x1b')
result.terminal.send('\x04')
result.terminal.send('\x03')
result.terminal.send('\x12')
result.terminal.send('\x0f')
result.terminal.send('/cancel')
result.terminal.send('\r')
expect(result.agent.cancelled).toContain('cancelled from terminal')
result.agent.status = 'idle'
for (const command of ['/help', '/reasoning', '/tools', '/redraw']) {
result.terminal.send(command)
result.terminal.send('\r')
await tick()
}
for (const command of ['/clear', '/cancel', '/wat']) {
result.terminal.send(command)
result.terminal.send('\r')
}
await tick()
result.terminal.send('draft')
result.terminal.send('\x03')
result.terminal.send('\x04')
await tick()
expect(result.terminal.output).toContain('Keyboard shortcuts')
expect(result.terminal.output).toContain('Reasoning blocks')
expect(result.terminal.output).toContain('Tool cards')
expect(result.terminal.output).toContain('already idle')
expect(result.terminal.output).toContain('Unknown command')
expect(result.exit).toHaveBeenCalledWith(0)
await result.controller.dispose()
await result.ctx.fiber.dispose()
const ctrlCExit = await setup()
ctrlCExit.terminal.send('\x03')
await tick()
expect(ctrlCExit.exit).toHaveBeenCalledWith(0)
await ctrlCExit.controller.dispose()
await ctrlCExit.ctx.fiber.dispose()
const disposedAgent = await setup()
disposedAgent.agent.status = 'disposed'
disposedAgent.terminal.send('late input')
disposedAgent.terminal.send('\r')
await tick()
expect(disposedAgent.terminal.output).toContain('is disposed')
await dispose(disposedAgent)
})
it('cancels before /exit while running and handles agent errors/disposal', async () => {
const result = await setup({ status: 'running' })
result.terminal.send('/exit')
result.terminal.send('\r')
await tick()
expect(result.agent.cancelled).toContain('terminal exit requested')
expect(result.exit).toHaveBeenCalledWith(0)
const events = await setup()
const unrelatedSession = events.ctx.sessions.create(SessionId('unrelated-session'))
const unrelatedAgent = { ...events.agent, id: AgentId('unrelated'), session: unrelatedSession }
unrelatedSession.append('todo/write', { todos: [{ content: 'hidden', status: 'pending' }] })
events.ctx.emit('agent/status', unrelatedAgent, 'running')
events.ctx.emit('agent/error', unrelatedAgent, 1, 1, new Error('hidden error'))
events.ctx.emit('agent/disposed', unrelatedAgent)
events.ctx.emit('agent/error', events.agent, 3, 2, new Error('live failure'))
events.session.append('turn/end', { turn: 3, reason: { kind: 'error', step: 2, message: 'live failure' } })
events.session.append('turn/end', { turn: 4, reason: { kind: 'error', step: 1, message: 'durable failure' } })
events.session.append('turn/end', { turn: 5, reason: { kind: 'aborted', reason: 'stopped' } })
events.session.append('turn/end', { turn: 6, reason: { kind: 'max-tokens' } })
events.session.append('turn/end', { turn: 7, reason: { kind: 'rejected', reason: 'policy' } })
events.session.append('turn/end', { turn: 8, reason: { kind: 'interrupted' } })
events.ctx.emit('agent/disposed', events.agent)
await tick()
expect(events.terminal.output).toContain('live failure')
expect(events.terminal.output).toContain('durable failure')
expect(events.terminal.output).toContain('stopped')
expect(events.terminal.output).toContain('output-token limit')
expect(events.terminal.output).toContain('Turn rejected')
expect(events.terminal.output).toContain('previous process ended')
expect(events.terminal.output).toContain('was disposed')
await dispose(events)
})
})
describe('tool cards and surface replay', () => {
const tools: Record<string, ToolDefinition> = {
bash: {
name: 'bash', description: '', parameters: {}, execute: async () => [],
presentCall: () => ({ card: 'terminal', title: 'printf hello', description: 'Run command', cwd: '/tmp' }),
presentResult: () => ({ card: 'terminal', output: 'hello\nworld\nthird', exitCode: 0 }),
},
signal: {
name: 'signal', description: '', parameters: {}, execute: async () => [],
presentCall: () => ({ card: 'terminal', title: 'sleep 10' }),
presentResult: () => ({ card: 'terminal', signal: 'SIGTERM' }),
},
edit: {
name: 'edit', description: '', parameters: {}, execute: async () => [],
presentCall: () => ({
card: 'diff',
title: 'Edit files',
diffs: [
{ path: 'a.txt', oldText: 'old', newText: 'new' },
{ path: 'b.txt', oldText: 'before', newText: 'after' },
],
}),
presentResult: () => ({ card: 'diff', diffs: [{ path: 'a.txt', oldText: null, newText: 'created' }] }),
},
generic: {
name: 'generic', description: '', parameters: {}, execute: async () => [],
presentCall: () => ({ card: 'generic', title: 'Inspect value', rawInput: { alpha: 1 } }),
presentResult: () => ({ card: 'generic', title: 'Inspected', content: [{ type: 'text', text: 'result text' }] }),
},
throwing: {
name: 'throwing', description: '', parameters: {}, execute: async () => [],
presentCall: () => { throw new Error('call presenter boom') },
presentResult: () => { throw new Error('result presenter boom') },
},
rawTerminal: {
name: 'rawTerminal', description: '', parameters: {}, execute: async () => [],
presentCall: () => ({ card: 'terminal', title: 'raw command' }),
},
undefinedViews: {
name: 'undefinedViews', description: '', parameters: {}, execute: async () => [],
presentCall: () => undefined,
presentResult: () => undefined,
},
empty: {
name: 'empty', description: '', parameters: {}, execute: async () => [],
presentCall: () => ({ card: 'generic', title: 'Empty card' }),
},
terminalResult: {
name: 'terminalResult', description: '', parameters: {}, execute: async () => [],
presentCall: () => ({ card: 'generic', title: 'Becomes terminal' }),
presentResult: () => ({ card: 'terminal', output: 'converted terminal' }),
},
symbolic: {
name: 'symbolic', description: '', parameters: {}, execute: async () => [],
presentCall: () => ({ card: 'generic', title: 'Symbol input', rawInput: Symbol('input') }),
},
}
it('uses terminal, diff, generic, fallback, and collapsed tool presentations', async () => {
const result = await setup({ tools, config: { maxToolOutputLines: 1 } })
const calls = [
['c1', 'bash', '{"command":"printf hello"}'],
['c2', 'signal', '{}'],
['c3', 'edit', '{}'],
['c4', 'generic', '{}'],
['c5', 'throwing', '{}'],
['c6', 'unknown', 'not-json'],
['c7', 'rawTerminal', '{"value":"raw"}'],
['c8', 'undefinedViews', '{"value":8}'],
['c10', 'empty', '{}'],
['c11', 'terminalResult', '{}'],
['c12', 'symbolic', '{}'],
] as const
appendAssistant(result.session, [
{ type: 'text', text: 'Calling tools' },
...calls.map(([id, name, args]) => ({
type: 'tool-call' as const, id: id as never, name, arguments: args,
})),
])
for (const [id, name, args] of calls) {
result.session.append('tool/call', { turn: 1, step: 0, callId: id as never, name, arguments: args })
}
await tick()
expect(result.terminal.output).toContain('$ raw command')
result.terminal.send('/reasoning')
result.terminal.send('\r')
await tick()
expect(result.terminal.output).toContain('call presenter boom')
expect(result.terminal.output).toContain('Symbol(input)')
result.session.append('tool/result', {
turn: 1, step: 0, callId: 'c1' as never, content: [{ type: 'text', text: 'raw bash' }], isError: false,
}, { surfaceOp: 'append' })
result.session.append('tool/result', {
turn: 1, step: 0, callId: 'c2' as never, content: [{ type: 'text', text: 'stopped' }], isError: true,
}, { surfaceOp: 'append' })
result.session.append('tool/result', {
turn: 1, step: 0, callId: 'c3' as never, content: [{ type: 'text', text: 'done' }], isError: false,
}, { surfaceOp: 'append' })
result.session.append('tool/result', {
turn: 1, step: 0, callId: 'c4' as never, content: [{ type: 'text', text: 'raw generic' }], isError: false,
}, { surfaceOp: 'append' })
result.session.append('tool/result', {
turn: 1, step: 0, callId: 'c5' as never, content: [{ type: 'text', text: 'raw throwing' }], isError: false,
meta: { value: 1 },
}, { surfaceOp: 'append' })
result.session.append('tool/result', {
turn: 1, step: 0, callId: 'c7' as never,
content: [
{ type: 'tool-call', id: 'inner' as never, name: 'inner', arguments: '{}' },
{ type: 'tool-result', toolCallId: 'inner' as never, content: [{ type: 'text', text: 'nested output' }] },
{ type: 'future-result' } as never,
],
isError: false,
}, { surfaceOp: 'append' })
result.session.append('tool/result', {
turn: 1, step: 0, callId: 'c8' as never, content: [{ type: 'text', text: '\nundefined presenter output\n\nkept tail\n' }], isError: false,
}, { surfaceOp: 'append' })
result.session.append('tool/result', {
turn: 1, step: 0, callId: 'c11' as never, content: [{ type: 'text', text: '\nconverted terminal\n\nfinished\n' }], isError: false,
}, { surfaceOp: 'append' })
result.session.append('tool/result', {
turn: 1, step: 0, callId: 'orphan' as never, content: [{ type: 'text', text: 'orphan result' }], isError: false,
}, { surfaceOp: 'append' })
await tick()
const output = result.terminal.output
expect(output).toContain('Run command')
expect(output).toContain('printf hello')
expect(output).toContain('more lines')
expect(output).toContain('SIGTERM')
expect(output).toContain('Edit files')
expect(output).toContain('Inspected')
expect(output).toContain('result text')
expect(output).toContain('Presenter failed')
expect(output).toContain('not-json')
expect(output).toContain('nested output')
expect(output).toContain('[future-result]')
expect(output).toContain('undefined presenter output')
expect(output).toContain('Empty card')
expect(output).toContain('converted terminal')
expect(output).toContain('orphan result')
result.terminal.send('/redraw')
result.terminal.send('\r')
await tick()
result.terminal.send('\x0f')
await tick()
expect(result.terminal.output).toContain('world')
expect(result.terminal.output).toContain('+ created')
await dispose(result)
})
it('rebuilds after a surface replacement and hides shadowed tool calls', async () => {
const result = await setup({ tools })
appendUser(result.session, 'old prompt')
const assistant = result.session.append('assistant/message', {
turn: 1,
step: 0,
content: [{ type: 'tool-call', id: 'old-call' as never, name: 'bash', arguments: '{}' }],
}, { surfaceOp: 'append' })
result.session.append('tool/call', {
turn: 1, step: 0, callId: 'old-call' as never, name: 'bash', arguments: '{}',
})
const toolResult = result.session.append('tool/result', {
turn: 1, step: 0, callId: 'old-call' as never, content: [{ type: 'text', text: 'old output' }], isError: false,
}, { surfaceOp: 'append' })
const start = result.session.surface.nodes[0]?.seq as number
result.session.append('context/message', {
content: [{ type: 'text', text: 'summary replacement' }],
source: { kind: 'plugin', plugin: 'compact' },
}, {
surfaceOp: { op: 'replace', start, end: toolResult.seq },
sourceEventSeqs: [start, assistant.seq, toolResult.seq],
})
await tick()
result.terminal.resize(89)
await tick()
const lastFullRender = result.terminal.output.slice(result.terminal.output.lastIndexOf('\x1b[2J'))
expect(lastFullRender).toContain('summary replacement')
expect(lastFullRender).not.toContain('old output')
await dispose(result)
})
})
describe('TUI user-interaction dialogs', () => {
it('answers single-select, multi-select, custom, and optionless questions', async () => {
const result = await setup({ config: { maxQuestionOptions: 1 } })
const single = result.ctx.userInteraction.ask({
questions: [{
id: 'mode', header: 'Mode', question: 'Choose a mode',
options: [{ label: 'Safe', description: 'Use checks' }, { label: 'Fast' }],
}],
})
await tick()
expect(result.terminal.output).toContain('Choose a mode')
expect(result.terminal.output).toContain('1/2')
result.terminal.send('\x1b[B')
result.terminal.send('\r')
await expect(single).resolves.toEqual({ answers: [{ id: 'mode', selected: ['Fast'] }] })
const multi = result.ctx.userInteraction.ask({
questions: [{ id: 'targets', question: 'Pick targets', multiSelect: true, options: [{ label: 'Code' }, { label: 'Docs' }] }],
})
await tick()
result.terminal.send(' ')
result.terminal.send('\x1b[B')
result.terminal.send(' ')
result.terminal.send('\r')
await expect(multi).resolves.toEqual({ answers: [{ id: 'targets', selected: ['Code', 'Docs'] }] })
const custom = result.ctx.userInteraction.ask({
questions: [{ id: 'other', question: 'Choose or type', options: [{ label: 'Default' }] }],
})
await tick()
result.terminal.send('c')
result.terminal.send('my choice')
result.terminal.send('\r')
await expect(custom).resolves.toEqual({ answers: [{ id: 'other', selected: [], custom: 'my choice' }] })
const free = result.ctx.userInteraction.ask({ questions: [{ id: 'note', question: 'Add a note' }] })
await tick()
result.terminal.send('\r')
await tick()
expect(result.terminal.output).toContain('Enter an answer before submitting')
result.terminal.send('ship it')
result.terminal.send('\r')
await expect(free).resolves.toEqual({ answers: [{ id: 'note', selected: [], custom: 'ship it' }] })
await dispose(result)
})
it('handles option wrapping, deselection errors, and returning from custom input', async () => {
const result = await setup({ config: { color: true } })
const single = result.ctx.userInteraction.ask({
questions: [{ id: 'single', question: 'Single options', options: [{ label: 'One' }, { label: 'Two' }] }],
})
const singleRejected = expect(single).rejects.toMatchObject({ code: 'ASK_ABORTED' })
await tick()
expect(result.terminal.output).toContain('Two')
result.terminal.send('\x03')
await singleRejected
const answer = result.ctx.userInteraction.ask({
questions: [{
id: 'options',
question: 'Exercise options',
multiSelect: true,
options: [{ label: 'One', description: 'first' }, { label: 'Two' }],
}],
})
const rejected = expect(answer).rejects.toMatchObject({ code: 'ASK_ABORTED' })
await tick()
result.terminal.send('\x1b[A')
result.terminal.send('\x1b[B')
result.terminal.send('\x1b[B')
result.terminal.send('\x1b[A')
result.terminal.send(' ')
await tick()
result.terminal.send('x')
result.terminal.send(' ')
result.terminal.send('\r')
await tick()
expect(result.terminal.output).toContain('Select at least one option')
result.terminal.send('c')
await tick()
result.terminal.send('\x1b')
await tick()
expect(result.terminal.output).toContain('Space toggle')
result.terminal.send('\x03')
await rejected
await dispose(result)
})
it('asks batches in order and rejects cancelled or aborted work', async () => {
const result = await setup()
const preAborted = new AbortController()
preAborted.abort()
await expect(result.ctx.userInteraction.ask({
questions: [{ id: 'pre-aborted', question: 'Already cancelled?' }],
signal: preAborted.signal,
})).rejects.toMatchObject({ code: 'ASK_ABORTED' })
const batch = result.ctx.userInteraction.ask({
questions: [
{ id: 'first', question: 'First?', options: [{ label: 'Yes' }] },
{ id: 'second', question: 'Second?' },
],
})
await tick()
result.terminal.send('\r')
await tick()
expect(result.terminal.output).toContain('Second?')
result.terminal.send('done')
result.terminal.send('\r')
await expect(batch).resolves.toEqual({ answers: [
{ id: 'first', selected: ['Yes'] },
{ id: 'second', selected: [], custom: 'done' },
] })
const cancelled = result.ctx.userInteraction.ask({ questions: [{ id: 'cancel', question: 'Cancel?' }] })
const cancelledExpectation = expect(cancelled).rejects.toMatchObject({ code: 'ASK_ABORTED' })
await tick()
result.terminal.send('\x1b')
await cancelledExpectation
const controller = new AbortController()
const active = result.ctx.userInteraction.ask({ questions: [{ id: 'active', question: 'Active?' }], signal: controller.signal })
const queuedController = new AbortController()
const queued = result.ctx.userInteraction.ask({ questions: [{ id: 'queued', question: 'Queued?' }], signal: queuedController.signal })
const activeExpectation = expect(active).rejects.toMatchObject({ code: 'ASK_ABORTED' })
const queuedExpectation = expect(queued).rejects.toMatchObject({ code: 'ASK_ABORTED' })
await tick()
queuedController.abort()
controller.abort()
await activeExpectation
await queuedExpectation
await dispose(result)
})
it('rejects active and queued dialogs on disposal', async () => {
const result = await setup()
const active = result.ctx.userInteraction.ask({ questions: [{ id: 'active', question: 'Active?' }] })
const queued = result.ctx.userInteraction.ask({ questions: [{ id: 'queued', question: 'Queued?' }] })
const activeExpectation = expect(active).rejects.toMatchObject({ code: 'ASK_ABORTED' })
const queuedExpectation = expect(queued).rejects.toMatchObject({ code: 'ASK_ABORTED' })
await tick()
await result.controller.dispose()
await activeExpectation
await queuedExpectation
await expect(result.ctx.userInteraction.ask({ questions: [{ id: 'late', question: 'Late?' }] }))
.rejects.toMatchObject({ code: 'NO_PROVIDER' })
await result.ctx.fiber.dispose()
})
})
describe('terminal mounting', () => {
it('starts immediately when the configured agent already exists', async () => {
const ctx = new Context()
await ctx.plugin(SessionStore)
await ctx.plugin(AgentRegistry)
await ctx.plugin(UserInteractionService)
ctx.provide('tools', { get: () => undefined } as never)
const session = ctx.sessions.create(SessionId('mounted-session'))
ctx.agents.register({
id: AgentId('main'), options: {}, session, status: 'idle', ctx,
send() {}, steer() {}, inject() {}, cancel() {}, whenIdle: () => Promise.resolve(),
})
const terminal = new FakeTerminal()
mountTui(ctx, { color: false }, { terminal, exit: vi.fn() })
await tick()
expect(terminal.started).toBe(1)
await ctx.fiber.dispose()
})
it('waits for its configured agent before starting the TUI', async () => {
const ctx = new Context()
await ctx.plugin(SessionStore)
await ctx.plugin(AgentRegistry)
await ctx.plugin(UserInteractionService)
ctx.provide('tools', { get: () => undefined } as never)
const terminal = new FakeTerminal()
mountTui(ctx, { agent: 'main', color: false }, { terminal, exit: vi.fn() })
expect(terminal.started).toBe(0)
const otherSession = ctx.sessions.create(SessionId('other-session'))
ctx.agents.register({
id: AgentId('other'), options: {}, session: otherSession, status: 'idle', ctx,
send() {}, steer() {}, inject() {}, cancel() {}, whenIdle: () => Promise.resolve(),
})
expect(terminal.started).toBe(0)
const session = ctx.sessions.create(SessionId('late-session'))
const agent = {
id: AgentId('main'), options: {}, session, status: 'idle', ctx,
send() {}, steer() {}, inject() {}, cancel() {}, whenIdle: () => Promise.resolve(),
} as Agent
ctx.agents.register(agent)
await tick()
expect(terminal.started).toBe(1)
await ctx.fiber.dispose()
})
it('rolls back providers, listeners, and terminal state when startup fails', async () => {
const ctx = new Context()
await ctx.plugin(SessionStore)
await ctx.plugin(AgentRegistry)
await ctx.plugin(UserInteractionService)
ctx.provide('tools', { get: () => undefined } as never)
const session = ctx.sessions.create(SessionId('failed-start-session'))
ctx.agents.register({
id: AgentId('main'), options: {}, session, status: 'running', ctx,
send() {}, steer() {}, inject() {}, cancel() {}, whenIdle: () => Promise.resolve(),
})
const terminal = new FakeTerminal()
terminal.start = () => { throw new Error('terminal startup failed') }
expect(() => createTuiChat(ctx, { color: false }, { terminal, exit: vi.fn() }))
.toThrow('terminal startup failed')
expect(terminal.stopped).toBe(1)
expect(terminal.progress).toEqual([false, true, false])
await expect(ctx.userInteraction.ask({ questions: [{ id: 'late', question: 'Late?' }] }))
.rejects.toMatchObject({ code: 'NO_PROVIDER' })
session.append('assistant/chunk', {
turn: 1,
step: 0,
chunk: { type: 'text-delta', index: 0, text: 'must not render' },
})
await tick()
expect(terminal.output).not.toContain('must not render')
await ctx.fiber.dispose()
})
it('throws when createTuiChat is called without the configured agent', async () => {
const ctx = new Context()
await ctx.plugin(AgentRegistry)
await ctx.plugin(UserInteractionService)
ctx.provide('tools', { get: () => undefined } as never)
const runtime: TuiRuntime = { terminal: new FakeTerminal(), exit: vi.fn() }
expect(() => createTuiChat(ctx, { agent: 'missing' }, runtime)).toThrow('is not running')
await ctx.fiber.dispose()
})
})

View File

@@ -0,0 +1,33 @@
{
"extends": "../../../tsconfig.base.json",
"compilerOptions": {
"rootDir": "src",
"outDir": "lib/types"
},
"include": [
"src"
],
"references": [
{
"path": "../../../vendor/cordis"
},
{
"path": "../../../vendor/schemastery"
},
{
"path": "../../core/agent"
},
{
"path": "../../core/session"
},
{
"path": "../../llm/llm"
},
{
"path": "../../core/tools"
},
{
"path": "../user-interaction"
}
]
}

View File

@@ -21,7 +21,7 @@ When an answer includes `custom`, `selected` is empty; custom text is an overrid
## Role
This is the interface package. Model-facing consumers such as `@deepseek-ai/dsh-tool-ask-user` depend on this seam; UI front doors such as the `stdio-agent` readline module and the `acp` bridge provide the provider. The loop stays unchanged: a tool call awaits a promise, and the tool result resumes the normal agent loop.
This is the interface package. Model-facing consumers such as `@deepseek-ai/dsh-tool-ask-user` depend on this seam; UI front doors such as the interactive `dsh-tui`, line-oriented `dsh-stdio`, and structured `dsh-acp` channels provide the provider. The loop stays unchanged: a tool call awaits a promise, and the tool result resumes the normal agent loop.
## Model Experience