Merge remote-tracking branch 'origin/master' into worktree/pr743-merge-20260727
# Conflicts: # examples/tui-agent/tests/snapshots/cordis-dynamic-toolchain/terminal.expected.txt # packages/ui/tui/tests/snapshots/cordis-tools-pending.expected.txt
This commit is contained in:
Tianyi Cui
@@ -5,7 +5,7 @@ import { fileURLToPath } from 'node:url'
|
||||
import { afterAll, describe, expect, it, vi } from 'vitest'
|
||||
import type { Context } from 'cordis'
|
||||
import { agentEvents } from '@deepseek-ai/dsh-agent'
|
||||
import { CallId, ReasoningEffortId, type ContentBlock } from '@deepseek-ai/dsh-llm'
|
||||
import { CallId, type ContentBlock } from '@deepseek-ai/dsh-llm'
|
||||
import type {} from '@deepseek-ai/dsh-llm-retry'
|
||||
import { SessionId, type JsonValue, type Session } from '@deepseek-ai/dsh-session'
|
||||
import SystemPrompt from '@deepseek-ai/dsh-system-prompt'
|
||||
@@ -27,6 +27,8 @@ const REFRESHING = process.env.DSH_SNAPSHOT === 'refresh'
|
||||
|
||||
const CHECKPOINTS = [
|
||||
'conversation-streaming',
|
||||
'shell-prompt-multiline',
|
||||
'step-timing-completed',
|
||||
'retry-scheduled',
|
||||
'retry-recovered',
|
||||
'retry-cancelled',
|
||||
@@ -40,12 +42,12 @@ const CHECKPOINTS = [
|
||||
'advanced-cards-expanded',
|
||||
'untrusted-controls',
|
||||
'question-dialog',
|
||||
'question-dialog-single-option',
|
||||
'question-dialog-validation',
|
||||
'surface-before-compaction',
|
||||
'surface-after-compaction-narrow',
|
||||
'surface-after-compaction-wide',
|
||||
'model-selector',
|
||||
'model-effort-switching',
|
||||
'model-switching',
|
||||
'errors-and-help',
|
||||
'disposed-terminal',
|
||||
@@ -104,7 +106,7 @@ async function setupSnapshot(
|
||||
cwd: options.cwd === undefined ? '/workspace/project' : options.cwd,
|
||||
config: Object.assign({
|
||||
welcome: 'Snapshot agent ready.',
|
||||
color: true,
|
||||
theme: { color: true },
|
||||
title: 'DSH snapshot',
|
||||
}, options.config),
|
||||
})
|
||||
@@ -195,13 +197,12 @@ const ADVANCED_CARD_TOOLS: Record<string, ToolDefinition> = {
|
||||
),
|
||||
edit: visualTool(
|
||||
'edit',
|
||||
() => ({ card: 'diff', title: 'Edit renderer', diffs: [{ path: 'src/view.ts', oldText: 'old line', newText: 'new line' }] }),
|
||||
() => ({ card: 'diff', title: 'Edit src/view.ts', diffs: [{ path: 'src/view.ts', oldText: 'old line', newText: 'new line' }] }),
|
||||
// The real edit/write tools produce exactly one diff whose path the title
|
||||
// already names, so the card omits the redundant per-file header.
|
||||
(): ToolResultView => ({
|
||||
card: 'diff',
|
||||
diffs: [
|
||||
{ path: 'src/view.ts', oldText: 'old line\nkeep', newText: 'new line\nkeep' },
|
||||
{ path: 'tests/view.spec.ts', oldText: null, newText: 'expect(screen).toMatchSnapshot()' },
|
||||
],
|
||||
diffs: [{ path: 'src/view.ts', oldText: 'old line\nkeep', newText: 'new line\nkeep' }],
|
||||
}),
|
||||
),
|
||||
subagent: visualTool('subagent', args => ({
|
||||
@@ -209,12 +210,19 @@ const ADVANCED_CARD_TOOLS: Record<string, ToolDefinition> = {
|
||||
title: 'Delegate renderer audit',
|
||||
rawInput: (args as { prompt: string }).prompt,
|
||||
})),
|
||||
task_output: visualTool('task_output', args => ({
|
||||
card: 'generic',
|
||||
kind: 'read',
|
||||
title: `Read output from background task ${(args as { task_id: string }).task_id}`,
|
||||
rawInput: (args as { task_id: string }).task_id,
|
||||
})),
|
||||
task_output: visualTool(
|
||||
'task_output',
|
||||
args => ({
|
||||
card: 'generic',
|
||||
kind: 'read',
|
||||
title: `Read output from background task ${(args as { task_id: string }).task_id}`,
|
||||
rawInput: (args as { task_id: string }).task_id,
|
||||
}),
|
||||
() => ({
|
||||
card: 'generic',
|
||||
content: [{ type: 'text', text: '```console\nstarted background task bash-5\n```' }],
|
||||
}),
|
||||
),
|
||||
skill: visualTool('skill', args => ({
|
||||
card: 'generic',
|
||||
kind: 'read',
|
||||
@@ -228,46 +236,64 @@ const DISPLAYED_CONTROL_PROBE = String.raw`\x1b]2;snapshot-controlled\x07\x09\x7
|
||||
|
||||
describe('TUI terminal-state snapshots', () => {
|
||||
it('pins an in-flight reasoning and Markdown stream', async () => {
|
||||
let clock = new Date(2026, 6, 21, 14, 30, 0).getTime()
|
||||
const nowSpy = vi.spyOn(Date, 'now').mockImplementation(() => clock)
|
||||
const harness = await setupSnapshot()
|
||||
// Freeze the loader's first animation interval so this semantic snapshot
|
||||
// cannot select a different spinner frame under scheduler contention.
|
||||
const frozenLoaderTimer = setInterval(() => {}, 60_000)
|
||||
const intervals = vi.spyOn(globalThis, 'setInterval').mockImplementationOnce(() => frozenLoaderTimer)
|
||||
try {
|
||||
await renderAfter(harness, () => {
|
||||
harness.agent.status = 'running'
|
||||
harness.ctx.emit('agent/status', harness.agent, 'running')
|
||||
appendUser(harness.session, 'Show the live update.')
|
||||
harness.session.append('assistant/chunk', {
|
||||
turn: 1,
|
||||
step: 1,
|
||||
chunk: { type: 'block-start', index: 0, blockType: 'reasoning' },
|
||||
})
|
||||
harness.session.append('assistant/chunk', {
|
||||
turn: 1,
|
||||
step: 1,
|
||||
chunk: { type: 'reasoning-delta', index: 0, text: 'Inspecting width and styles.' },
|
||||
})
|
||||
harness.session.append('assistant/chunk', {
|
||||
turn: 1,
|
||||
step: 1,
|
||||
chunk: { type: 'block-start', index: 1, blockType: 'text' },
|
||||
})
|
||||
harness.session.append('assistant/chunk', {
|
||||
turn: 1,
|
||||
step: 1,
|
||||
chunk: { type: 'text-delta', index: 1, text: 'Streaming **visible state**…' },
|
||||
})
|
||||
await renderAfter(harness, () => {
|
||||
harness.agent.status = 'running'
|
||||
harness.ctx.emit('agent/status', harness.agent, 'running')
|
||||
appendUser(harness.session, 'Show the live update.')
|
||||
clock += 1_000
|
||||
harness.session.append('assistant/chunk', {
|
||||
turn: 1,
|
||||
step: 1,
|
||||
chunk: { type: 'block-start', index: 0, blockType: 'reasoning' },
|
||||
})
|
||||
const loaderIntervalMs = intervals.mock.calls[0]?.[1]
|
||||
if (typeof loaderIntervalMs !== 'number') throw new Error('TUI loader did not register an animation interval')
|
||||
await new Promise(resolve => setTimeout(resolve, loaderIntervalMs + 5))
|
||||
await checkpoint('conversation-streaming', harness.terminal)
|
||||
} finally {
|
||||
intervals.mockRestore()
|
||||
clearInterval(frozenLoaderTimer)
|
||||
await disposeSnapshot(harness)
|
||||
}
|
||||
harness.session.append('assistant/chunk', {
|
||||
turn: 1,
|
||||
step: 1,
|
||||
chunk: { type: 'reasoning-delta', index: 0, text: 'Inspecting width and styles.' },
|
||||
})
|
||||
clock += 2_000
|
||||
harness.session.append('assistant/chunk', {
|
||||
turn: 1,
|
||||
step: 1,
|
||||
chunk: { type: 'block-start', index: 1, blockType: 'text' },
|
||||
})
|
||||
harness.session.append('assistant/chunk', {
|
||||
turn: 1,
|
||||
step: 1,
|
||||
chunk: { type: 'text-delta', index: 1, text: 'Streaming **visible state**…\n\n```ts\nconst visible = true\n```' },
|
||||
})
|
||||
})
|
||||
await checkpoint('conversation-streaming', harness.terminal)
|
||||
await disposeSnapshot(harness)
|
||||
nowSpy.mockRestore()
|
||||
})
|
||||
|
||||
it('pins a completed step timing summary', async () => {
|
||||
let clock = new Date(2026, 6, 21, 14, 32, 6).getTime()
|
||||
const nowSpy = vi.spyOn(Date, 'now').mockImplementation(() => clock)
|
||||
const harness = await setupSnapshot()
|
||||
await renderAfter(harness, () => {
|
||||
clock += 1_000
|
||||
harness.session.append('assistant/chunk', {
|
||||
turn: 1,
|
||||
step: 1,
|
||||
chunk: { type: 'reasoning-delta', index: 0, text: 'Checking the result.' },
|
||||
})
|
||||
clock += 2_000
|
||||
harness.session.append('assistant/chunk', {
|
||||
turn: 1,
|
||||
step: 1,
|
||||
chunk: { type: 'text-delta', index: 1, text: 'The result is ready.' },
|
||||
})
|
||||
clock += 3_000
|
||||
harness.session.append('step/end', { turn: 1, step: 1 })
|
||||
})
|
||||
await checkpoint('step-timing-completed', harness.terminal, { includeScrollback: true })
|
||||
nowSpy.mockRestore()
|
||||
await disposeSnapshot(harness)
|
||||
})
|
||||
|
||||
it('pins failed-stream retraction, scheduled retry, and eventual success', async () => {
|
||||
@@ -293,15 +319,13 @@ describe('TUI terminal-state snapshots', () => {
|
||||
})
|
||||
await checkpoint('retry-scheduled', harness.terminal, { includeScrollback: true })
|
||||
|
||||
await renderAfter(harness, () => {
|
||||
harness.session.append('assistant/message', {
|
||||
turn: 1,
|
||||
step: 2,
|
||||
provenance: { provider: 'mock', model: 'deepseek-v4-flash' },
|
||||
content: [{ type: 'text', text: 'Recovered on the next bounded attempt.' }],
|
||||
}, { surfaceOp: 'append' })
|
||||
harness.session.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
|
||||
})
|
||||
harness.session.append('assistant/message', {
|
||||
turn: 1,
|
||||
step: 2,
|
||||
provenance: { provider: 'mock', model: 'deepseek-v4-flash' },
|
||||
content: [{ type: 'text', text: 'Recovered on the next bounded attempt.' }],
|
||||
}, { surfaceOp: 'append' })
|
||||
harness.session.append('turn/end', { turn: 1, reason: { kind: 'completed' } })
|
||||
await checkpoint('retry-recovered', harness.terminal, { includeScrollback: true })
|
||||
await disposeSnapshot(harness)
|
||||
})
|
||||
@@ -352,7 +376,7 @@ describe('TUI terminal-state snapshots', () => {
|
||||
})
|
||||
|
||||
it('paints the startup banner product name in the DeepSeek brand gradient on truecolor terminals', async () => {
|
||||
const harness = await setupSnapshot({ config: { truecolor: true } })
|
||||
const harness = await setupSnapshot({ config: { theme: { truecolor: true } } })
|
||||
await checkpoint('banner-gradient', harness.terminal, {}, true)
|
||||
await disposeSnapshot(harness)
|
||||
})
|
||||
@@ -457,6 +481,7 @@ describe('TUI terminal-state snapshots', () => {
|
||||
})
|
||||
|
||||
it('renders terminal controls as inert text across transcripts, tools, dialogs, diagnostics, and title', async () => {
|
||||
const nowSpy = vi.spyOn(Date, 'now').mockReturnValue(new Date(2026, 6, 21, 15, 0, 0).getTime())
|
||||
const tools = {
|
||||
unsafe: visualTool(
|
||||
'unsafe',
|
||||
@@ -518,14 +543,13 @@ describe('TUI terminal-state snapshots', () => {
|
||||
})
|
||||
const rejected = expect(answer).rejects.toMatchObject({ code: 'ASK_ABORTED' })
|
||||
await harness.terminal.waitForFrame(beforeQuestion)
|
||||
await renderAfter(harness, () => {
|
||||
agentEvents(harness.ctx, harness.agent).emit('agent/error', 8, 3, new Error(`Unsafe live error ${CONTROL_PROBE}`))
|
||||
})
|
||||
agentEvents(harness.ctx, harness.agent).emit('agent/error', 8, 3, new Error(`Unsafe live error ${CONTROL_PROBE}`))
|
||||
await checkpoint('untrusted-controls', harness.terminal, { includeScrollback: true })
|
||||
|
||||
controller.abort()
|
||||
await rejected
|
||||
await disposeSnapshot(harness)
|
||||
nowSpy.mockRestore()
|
||||
})
|
||||
|
||||
it('pins a constrained multi-select question and its validation state', async () => {
|
||||
@@ -568,7 +592,37 @@ describe('TUI terminal-state snapshots', () => {
|
||||
await disposeSnapshot(harness)
|
||||
})
|
||||
|
||||
it('pins a single-option question', async () => {
|
||||
const harness = await setupSnapshot({
|
||||
config: {
|
||||
questionDialogWidth: 200,
|
||||
questionDialogMaxHeight: 16,
|
||||
},
|
||||
}, { columns: 56, rows: 20 })
|
||||
const controller = new AbortController()
|
||||
const beforeQuestion = harness.terminal.frames
|
||||
const answer = harness.ctx.userInteraction.ask({
|
||||
questions: [{
|
||||
id: 'confirm',
|
||||
header: 'Confirm',
|
||||
question: 'Continue with this change?',
|
||||
options: [{ label: 'Proceed', description: 'Apply the proposed change' }],
|
||||
}],
|
||||
signal: controller.signal,
|
||||
})
|
||||
const rejected = expect(answer).rejects.toMatchObject({ code: 'ASK_ABORTED' })
|
||||
await harness.terminal.waitForFrame(beforeQuestion)
|
||||
await checkpoint('question-dialog-single-option', harness.terminal)
|
||||
controller.abort()
|
||||
await rejected
|
||||
await disposeSnapshot(harness)
|
||||
})
|
||||
|
||||
it('pins compaction surface replacement and narrow-to-wide reflow', async () => {
|
||||
// Freeze the clock: the timing header hides zero-duration buckets, so a
|
||||
// real-clock millisecond tick between the fixture appends and the render
|
||||
// would flip `Tools 0.0s` in and out of the pinned header.
|
||||
const nowSpy = vi.spyOn(Date, 'now').mockReturnValue(new Date(2026, 6, 21, 14, 40, 0).getTime())
|
||||
let replacementStart = 0
|
||||
let replacementEnd = 0
|
||||
let replacementSources: number[] = []
|
||||
@@ -602,8 +656,11 @@ describe('TUI terminal-state snapshots', () => {
|
||||
|
||||
await renderAfter(harness, () => {
|
||||
harness.session.append('user/message', {
|
||||
content: [{ type: 'text', text: 'Compacted summary: the prior command completed and its details were retired from the active surface.' }],
|
||||
source: { kind: 'plugin', plugin: 'compact' },
|
||||
content: [{
|
||||
type: 'text',
|
||||
text: '<system-reminder>\nAdditional instructions from: nested/AGENTS.md\n\nRender workspace context XML clearly.\n</system-reminder>',
|
||||
}],
|
||||
source: { kind: 'plugin', plugin: 'workspace-context' },
|
||||
}, {
|
||||
surfaceOp: { op: 'replace', start: replacementStart, end: replacementEnd },
|
||||
sourceEventSeqs: replacementSources,
|
||||
@@ -615,9 +672,22 @@ describe('TUI terminal-state snapshots', () => {
|
||||
await renderAfter(harness, () => { harness.terminal.resize(104, 30) })
|
||||
await checkpoint('surface-after-compaction-wide', harness.terminal, { includeScrollback: true })
|
||||
await disposeSnapshot(harness)
|
||||
nowSpy.mockRestore()
|
||||
})
|
||||
|
||||
it('pins wrapped and explicit multiline shell-prompt input', async () => {
|
||||
const harness = await setupSnapshot({}, { columns: 44, rows: 18 })
|
||||
await renderAfter(harness, () => {
|
||||
harness.terminal.send('Explain this implementation with enough detail to wrap across multiple full-width continuation rows without leaving a prompt-sized gap at the right edge.')
|
||||
harness.terminal.send('\x1b[13;2u')
|
||||
harness.terminal.send('Then suggest a simpler version.')
|
||||
})
|
||||
await checkpoint('shell-prompt-multiline', harness.terminal)
|
||||
await disposeSnapshot(harness)
|
||||
})
|
||||
|
||||
it('pins help, unknown commands, live errors, turn failures, and terminal restoration', async () => {
|
||||
const nowSpy = vi.spyOn(Date, 'now').mockReturnValue(new Date(2026, 6, 21, 15, 5, 0).getTime())
|
||||
const harness = await setupSnapshot({}, { columns: 92, rows: 32 })
|
||||
await renderAfter(harness, () => {
|
||||
harness.terminal.send('/help')
|
||||
@@ -635,6 +705,12 @@ describe('TUI terminal-state snapshots', () => {
|
||||
turn: 2,
|
||||
reason: { kind: 'interrupted' },
|
||||
})
|
||||
harness.session.append('turn/start', { turn: 3, trigger: { kind: 'message', source: { kind: 'user' } } })
|
||||
harness.session.append('turn/end', { turn: 3, reason: { kind: 'disposed' } })
|
||||
harness.session.append('turn/start', { turn: 4, trigger: { kind: 'message', source: { kind: 'user' } } })
|
||||
// A merge-extensible turn-end kind unknown to the TUI still surfaces its
|
||||
// name so the agent never stops without a visible reason.
|
||||
harness.session.append('turn/end', { turn: 4, reason: { kind: 'plugin-policy' } as never })
|
||||
})
|
||||
await checkpoint('errors-and-help', harness.terminal, { includeScrollback: true })
|
||||
|
||||
@@ -643,31 +719,11 @@ describe('TUI terminal-state snapshots', () => {
|
||||
await checkpoint('disposed-terminal', harness.terminal, { includeScrollback: true })
|
||||
await harness.ctx.fiber.dispose()
|
||||
await harness.terminal.dispose()
|
||||
nowSpy.mockRestore()
|
||||
})
|
||||
|
||||
it('pins the model selector, effort cycling, and provider-default selection', async () => {
|
||||
const harness = await setupSnapshot({
|
||||
catalog: {
|
||||
providers: [{ id: 'deepseek', name: 'DeepSeek' }],
|
||||
models: [
|
||||
{ provider: 'deepseek', id: 'deepseek-v4-flash', name: 'DeepSeek V4 Flash' },
|
||||
{ provider: 'deepseek', id: 'deepseek-v4-pro', name: 'DeepSeek V4 Pro' },
|
||||
],
|
||||
resolveModelInfo: (_provider, model) => Promise.resolve({
|
||||
context: { contextWindow: 128_000 },
|
||||
reasoning: {
|
||||
efforts: [
|
||||
{ id: ReasoningEffortId('off'), name: 'Off' },
|
||||
{ id: ReasoningEffortId('high'), name: 'High' },
|
||||
{ id: ReasoningEffortId('max'), name: 'Max' },
|
||||
],
|
||||
...model === 'deepseek-v4-flash'
|
||||
? { defaultEffort: ReasoningEffortId('high') }
|
||||
: {},
|
||||
},
|
||||
}),
|
||||
},
|
||||
}, { columns: 92, rows: 32 })
|
||||
it('pins the model selector and selection notice', async () => {
|
||||
const harness = await setupSnapshot({}, { columns: 92, rows: 32 })
|
||||
await renderAfter(harness, () => {
|
||||
harness.terminal.send('/model')
|
||||
harness.terminal.send('\r')
|
||||
@@ -675,13 +731,6 @@ describe('TUI terminal-state snapshots', () => {
|
||||
await checkpoint('model-selector', harness.terminal, { includeScrollback: true })
|
||||
await renderAfter(harness, () => {
|
||||
harness.terminal.send('\x1b[B')
|
||||
harness.terminal.send('\x1b[Z')
|
||||
harness.terminal.send('\x1b[Z')
|
||||
harness.terminal.send('\x1b[Z')
|
||||
harness.terminal.send('\x1b[Z')
|
||||
})
|
||||
await checkpoint('model-effort-switching', harness.terminal, { includeScrollback: true })
|
||||
await renderAfter(harness, () => {
|
||||
harness.terminal.send('\r')
|
||||
})
|
||||
await checkpoint('model-switching', harness.terminal, { includeScrollback: true })
|
||||
@@ -727,6 +776,22 @@ describe('TUI terminal-state snapshots', () => {
|
||||
contextWindow: 128_000,
|
||||
contextTokens: 42_000,
|
||||
agentOptions: { provider: 'deepseek', model: 'deepseek-v4-pro' },
|
||||
tools: {
|
||||
read: {
|
||||
name: 'read',
|
||||
description: 'Read a file',
|
||||
parameters: {},
|
||||
output: { schema: { type: 'null' }, render: () => [] },
|
||||
execute: async () => null,
|
||||
},
|
||||
write: {
|
||||
name: 'write',
|
||||
description: 'Write a file',
|
||||
parameters: {},
|
||||
output: { schema: { type: 'null' }, render: () => [] },
|
||||
execute: async () => null,
|
||||
},
|
||||
},
|
||||
beforeMount(session) {
|
||||
appendUser(session, 'inspect this session')
|
||||
appendAssistant(session, [{ type: 'text', text: 'Session inspected.' }], {
|
||||
|
||||
Reference in New Issue
Block a user