Merge branch 'master' into worktree/web-skill-tool-row

apps/cli/composition.md

@@ -40,6 +40,8 @@ flowchart LR
   cfg --> plugin_dsh_base_session_persistence_jsonl
   plugin_dsh_base_session_query_sqlite["session-query-sqlite<br/>@deepseek-ai/dsh-session-query-sqlite"]
   cfg --> plugin_dsh_base_session_query_sqlite
+  plugin_dsh_base_session_projection["session-projection<br/>@deepseek-ai/dsh-session-projection"]
+  cfg --> plugin_dsh_base_session_projection
   plugin_dsh_base_telemetry_otel["telemetry-otel<br/>@deepseek-ai/dsh-session-telemetry-otel"]
   cfg --> plugin_dsh_base_telemetry_otel
   plugin_dsh_base_subprocess["subprocess<br/>@deepseek-ai/dsh-subprocess-local"]
@@ -166,6 +168,7 @@ flowchart LR
 | `llm-pi-ai` | `@deepseek-ai/dsh-llm-pi-ai` |
 | `session-persistence-jsonl` | `@deepseek-ai/dsh-session-persistence-jsonl` |
 | `session-query-sqlite` | `@deepseek-ai/dsh-session-query-sqlite` |
+| `session-projection` | `@deepseek-ai/dsh-session-projection` |
 | `telemetry-otel` | `@deepseek-ai/dsh-session-telemetry-otel` |
 | `subprocess` | `@deepseek-ai/dsh-subprocess-local` |
 | `sandbox` | `@deepseek-ai/dsh-sandbox-local` |

apps/cli/config/core-web.cordis.yml

@@ -1,6 +1,27 @@
-# Opt-in two-tool profile over the shipped Web composition. The default native
-# model surface is exactly persistent `bash` plus `str_replace_editor`; the
-# Web host, browser shell, workspace, persistence, and permission stack remain.
+# Opt-in Web shell for the RL core agent contract. The model receives exactly
+# the configured persona plus the native `bash` and `str_replace_editor`
+# schemas; the Web host, browser shell, persistence, and permission stack stay.
+
+# Match the Claude SWE-compatible RL core prompt. Disabling the Web runtime's
+# surface context removes its GUI orientation, managed shell variables, and the
+# launcher's source-checkout section through one configuration contract.
+# Workspace instructions are model-visible user context rather than a system
+# section, but RL core disables them as part of the same prompt contract.
+- id: system-prompt
+  config:
+    includeHarnessIdentity: false
+    persona: !!js process.env.DSH_SYSTEM_PROMPT ?? 'You are a helpful software engineer assistant.'
+
+- id: web-runtime
+  config:
+    surfaceContext: false
+
+- id: workspace-context
+  disabled: true
+
+- id: tools
+  config:
+    mode: native
 
 # Disable every model-facing consumer in the base/Web tree. plan-mode owns the
 # always-registered exit_plan_mode tool even while the session is not planning.

apps/cli/reference/README.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 apps/cli/reference/README.md
-README.md: 62275139809364d80296433e804ade3d447f37e6
-README.zh.md: e508caaf05622c0c177eaacd070978368ff84af3
+README.md: 8b8a0e7dbebafedd6a4f8d988adb3fd11c7bd026
+README.zh.md: d1d6d5a594596a8be5db30021163f0fcea4a95bf

apps/cli/reference/README.md

@@ -53,7 +53,9 @@ All modes treat the invoking directory as the default workspace root, load appli
 
 New sessions default to the `workspace-write` permission preset. Bash and filesystem mutations are restricted to the session workspace and platform temporary roots; reads, network access, and process visibility are not confined. `DSH_PERMISSION_MODE` changes the process fallback. Stored General-settings permissions affect later Web sessions, not an already-open one.
 
-`DSH_TOOLS_MODE` selects `native`, `code`, or `both` for the process; another value fails at boot. [`config/core-web.cordis.yml`](../config/core-web.cordis.yml) is an optional `--patch` overlay that reduces the native model surface to persistent `bash` and `str_replace_editor` while retaining the shipped host, browser, workspace, persistence, and permission composition.
+`DSH_TOOLS_MODE` selects `native`, `code`, or `both` for the process; another value fails at boot. [`config/core-web.cordis.yml`](../config/core-web.cordis.yml) is an optional RL-compatible `--patch` overlay that pins native mode, renders only `DSH_SYSTEM_PROMPT` or `You are a helpful software engineer assistant.` as the system prompt, disables Workspace instructions and every Web runtime prompt contribution, and exposes only persistent `bash` and `str_replace_editor` while retaining the shipped host, browser, workspace, persistence, and permission composition.
+
+`DSH_SYSTEM_PROMPT` is passed as the system-prompt [`persona`](../../../packages/core/system-prompt/README.md#config): complete `{{…}}` groups use that contract's strict variable interpolation and have no literal-brace escape; any set value, including an empty string, is authoritative and an empty value therefore removes the system prompt, while only an unset variable selects the fallback.
 
 ## Shared deployment behavior
 

apps/cli/reference/README.zh.md

@@ -53,7 +53,9 @@ dsh web --dump-config
 
 新会话默认使用 `workspace-write` 权限预设。Bash 和文件系统修改仅限于会话 workspace 与平台临时根目录；读取、网络访问和进程可见性不受限制。`DSH_PERMISSION_MODE` 更改进程后备值。General settings 中存储的权限影响后续 Web 会话，不改变已打开的会话。
 
-`DSH_TOOLS_MODE` 为进程选择 `native`、`code` 或 `both`；其他值会导致启动失败。[`config/core-web.cordis.yml`](../config/core-web.cordis.yml) 是可选的 `--patch` overlay：它在保留随附宿主、浏览器、workspace、持久化和权限组合的同时，把原生模型 surface 缩减为持久 `bash` 和 `str_replace_editor`。
+`DSH_TOOLS_MODE` 为进程选择 `native`、`code` 或 `both`；其他值会导致启动失败。[`config/core-web.cordis.yml`](../config/core-web.cordis.yml) 是可选的 RL 兼容 `--patch` overlay：它固定使用 `native` 模式，仅将 `DSH_SYSTEM_PROMPT` 或 `You are a helpful software engineer assistant.` 渲染为系统提示词，禁用 Workspace 指令与所有 Web 运行时提示词贡献，并且在保留随附宿主、浏览器、workspace、持久化和权限组合的同时，仅暴露持久 `bash` 和 `str_replace_editor`。
+
+`DSH_SYSTEM_PROMPT` 会传给系统提示词的 [`persona`](../../../packages/core/system-prompt/README.md#config)：完整的 `{{…}}` 分组遵循该契约的严格变量插值规则，且无法转义为字面花括号；任何已设置的值（包括空字符串）都具有权威性，因此空值会移除系统提示词，只有未设置该变量时才会选择后备值。
 
 ## 共享部署行为
 

apps/cli/src/profile-boot.ts

@@ -95,6 +95,9 @@ export function prepareProfile(name: string, userLayer = true): Profile {
   return profile
 }
 
+/** Read-only row index of a profile composition before launcher flag patches. */
+export type ProfileRows = ReadonlyMap<string, { name?: string; config?: unknown }>
+
 /** One profile's patch layers (application order) and the row index of its pre-flag composition. */
 interface ComposedProfile {
   profile: Profile
@@ -109,7 +112,7 @@ interface ComposedProfile {
    * for flag merges and row checks. Flag patches must not insert rows the
    * launcher consults here (they only override values and insert dev glue).
    */
-  rows: Map<string, { name?: string; config?: unknown }>
+  rows: ProfileRows
 }
 
 /** The full patch stack of one composed profile, in application order. */
@@ -155,11 +158,11 @@ export interface RunProfileOptions {
   /** `--patch` overlay paths, in argv order. */
   patchFiles: readonly string[]
   /** Launcher hook turning the pre-flag composed rows into flag patches (the web alias's flag family). */
-  deriveFlagPatches?: (rows: ComposedProfile['rows']) => PatchOptions[]
+  deriveFlagPatches?: (rows: ProfileRows) => PatchOptions[]
   /** One-shot task text; requires the composition to mount the headless runner row. */
   task?: string
   /** Surface setup registered after Loader installation and before any config-tree entry mounts. */
-  prepare?: (ctx: Context) => Promise<void> | void
+  prepare?: (ctx: Context, rows: ProfileRows) => Promise<void> | void
 }
 
 /**
@@ -231,7 +234,7 @@ export async function runProfile(options: RunProfileOptions): Promise<{ ctx: Con
       }
       hostCtx.provide('headlessIo', io)
     }
-    await options.prepare?.(hostCtx)
+    await options.prepare?.(hostCtx, composed.rows)
   })
   app.current = ctx
   // A surface can dispose the whole tree while startup was still in flight

apps/cli/src/web.ts

@@ -13,7 +13,7 @@ import { fileURLToPath } from 'node:url'
 import type { Context } from 'cordis'
 import type { PatchOptions } from '@cordisjs/plugin-include'
 import { addHarnessSourceSection } from '@deepseek-ai/dsh-app-boot'
-import { runProfile } from './profile-boot.ts'
+import { runProfile, type ProfileRows } from './profile-boot.ts'
 
 const SOURCE_ROOT = fileURLToPath(new URL('../../..', import.meta.url))
 
@@ -71,7 +71,7 @@ export interface WebFlags {
  * @returns the flag patch list, in application order.
  */
 function deriveWebFlagPatches(
-  rows: Map<string, { name?: string; config?: unknown }>,
+  rows: ProfileRows,
   flags: WebFlags,
 ): PatchOptions[] {
   const overrides = new Map<string, Record<string, unknown>>()
@@ -105,6 +105,18 @@ function deriveWebFlagPatches(
   return patches
 }
 
+/**
+ * Whether the composed Web runtime keeps its model- and shell-visible surface
+ * context. The bundle schema defaults the field to true, so only an explicit
+ * false suppresses both the bundle contributions and the launcher-owned
+ * source-checkout section.
+ * @param rows - the composed Web profile rows before launcher flag patches.
+ * @returns true unless the web-runtime row explicitly disables surface context.
+ */
+export function webSurfaceContextEnabled(rows: ProfileRows): boolean {
+  return (rows.get('web-runtime')?.config as { surfaceContext?: boolean } | undefined)?.surfaceContext !== false
+}
+
 /**
  * Serve the browser UI from the web profile. Host/port/workspace-root flags
  * are passed through only when given (absent, the composed profile values
@@ -118,7 +130,8 @@ export async function runWeb(flags: WebFlags): Promise<void> {
     profile: 'web',
     patchFiles: flags.patches,
     deriveFlagPatches: rows => deriveWebFlagPatches(rows, flags),
-    prepare: (ctx: Context) => {
+    prepare: (ctx: Context, rows: ProfileRows) => {
+      if (!webSurfaceContextEnabled(rows)) return
       ctx.inject(['systemPrompt'], (promptCtx) => {
         addHarnessSourceSection(promptCtx, SOURCE_ROOT)
       })

apps/cli/tests/built-bin.e2e.ts

@@ -8,6 +8,7 @@ import { afterEach, beforeEach, describe, expect, it } from 'vitest'
 /** Published-entry acceptance for argument errors, profile lifecycle, and boot-free config dumps. */
 const repoRoot = fileURLToPath(new URL('../../../', import.meta.url))
 const dshBin = join(repoRoot, 'apps/cli/lib/bin.js')
+const coreWebOverlay = fileURLToPath(new URL('../config/core-web.cordis.yml', import.meta.url))
 const invalidProvider = fileURLToPath(new URL('./fixtures/invalid-provider.cordis.yml', import.meta.url))
 
 async function runBuiltBin(
@@ -367,5 +368,16 @@ describe.skipIf(!existsSync(dshBin))('dsh BUILT bin (node lib/bin.js, no tsx)',
       expect(stdout).toContain(`patched by ${profilePatch}, ${overlay}`)
       expect(stderr).toContain('patch: entry "absent-row" not found')
     }, 30_000)
+
+    it('shows the RL Web patch disabling runtime surface context', async () => {
+      const { stdout, code, stderr } = await runBuiltBin(
+        ['web', '--patch', coreWebOverlay, '--dump-config'],
+        { DSH_HOME: home },
+      )
+      expect(code).toBe(0)
+      expect(stderr).toBe('')
+      expect(stdout).toContain("name: '@deepseek-ai/dsh-web-app'")
+      expect(stdout).toContain('surfaceContext: false')
+    }, 30_000)
   })
 })

apps/cli/tests/trusted-hosts.spec.ts

@@ -1,7 +1,7 @@
 /** Single-sample LAN-trust resolution for the /api browser-trust fence (`resolveLanTrust`). */
 
 import { describe, expect, it, vi } from 'vitest'
-import { resolveLanTrust } from '../src/web.ts'
+import { resolveLanTrust, webSurfaceContextEnabled } from '../src/web.ts'
 
 vi.mock('node:os', () => ({
   networkInterfaces: () => ({
@@ -31,3 +31,15 @@ describe('resolveLanTrust', () => {
     expect(resolveLanTrust(undefined, ['lab.internal'])).toEqual({ lanAddresses: [], trustedHosts: ['lab.internal'] })
   })
 })
+
+describe('webSurfaceContextEnabled', () => {
+  it('defaults to enabled and honors an explicit complete-prompt disable', () => {
+    expect(webSurfaceContextEnabled(new Map())).toBe(true)
+    expect(webSurfaceContextEnabled(new Map([
+      ['web-runtime', { config: { mode: 'production' } }],
+    ]))).toBe(true)
+    expect(webSurfaceContextEnabled(new Map([
+      ['web-runtime', { config: { surfaceContext: false } }],
+    ]))).toBe(false)
+  })
+})

apps/web/tests/conversation-column-overflow.e2e.ts

@@ -0,0 +1,339 @@
+// Web e2e scenario: the conversation column scrolls on one axis only, as the
+// browser actually lays it out. The reported symptom was a horizontal
+// scrollbar under the whole center column once the window (or the sidebar
+// drag) narrowed it — the hero's decorative backdrop ellipse bleeding past the
+// column and becoming user-scrollable.
+//
+// The bleed is by construction and stays: `.heroGlow` is sized 1051/776 of the
+// hero box (ConversationRoot.module.css) so the blur scales with the input
+// card. What changed is the scroll container: `[data-conversation-scroll]`
+// scrolls vertically, and a box that scrolls in one axis computes the other
+// axis's initial `visible` to `auto`, so the bleed came back as a bar. The
+// fix states `overflow-x: hidden` there.
+//
+// Only a real engine reports that pair — the bleed and the resulting scroll
+// range — so the scenario sweeps viewport widths that bracket the glow's
+// width and asserts both at each stop. Asserting no horizontal scroll alone
+// would go vacuous the moment the glow stopped bleeding for an unrelated
+// reason, which is why each stop also records whether it bleeds; the wide stop
+// is the control where it does not.
+//
+// Zero model calls: the hero is the boot state, so nothing is seeded and no
+// replay row mounts. A stray stream would fail loud with NO_ADAPTER.
+import { fileURLToPath } from 'node:url'
+import { join } from 'node:path'
+import type { Browser, Page } from 'playwright'
+import { chromium } from 'playwright'
+import { afterAll, beforeAll, describe, expect, it, onTestFailed } from 'vitest'
+import {
+  assertFixtureInventory, compareOrRefreshGolden, launchWebScaffold, watchConsole, webSnapshotMode,
+  type WebScaffold,
+} from './scaffold.ts'
+import { newEnglishPage, saveFailureShot } from './support.ts'
+
+const SNAPSHOT_DIR = fileURLToPath(new URL('./snapshots/conversation-column-overflow', import.meta.url))
+/**
+ * Committed golden of the one-axis relation at every stop. It records
+ * relations and booleans, never absolute coordinates: the column width follows
+ * the viewport and the sidebar, and a golden carrying pixels would document the
+ * platform instead of the change.
+ */
+const GEOMETRY_EXPECTED = join(SNAPSHOT_DIR, 'geometry.expected.md')
+const MODE = webSnapshotMode()
+/** Narrow sweep stop where the mutation control retains overflow across scrollbar implementations. */
+const CONTROL_VIEWPORT = 600
+/**
+ * Viewport widths bracketing the glow: the narrow stops retain the reported
+ * bleed while the widest stop proves the relation can also be false.
+ */
+const WIDTHS = [1680, 1200, 1000, 800, CONTROL_VIEWPORT]
+/** Element id of the mutation control's injected sheet, so the test can take it back out. */
+const CONTROL_STYLE_ID = 'dsh-column-overflow-control'
+/** Horizontal wheel delta per gesture; must exceed the widest bleed the sweep can produce. */
+const WHEEL_DELTA = 300
+
+/** One viewport stop: whether the glow bleeds past the column, and whether that bleed scrolls. */
+interface ColumnMetrics {
+  /** Viewport width the stop was measured at. */
+  width: number
+  /** The column's content width. Not committed to the golden — it is what settles after a resize, and what the sweep waits on. */
+  columnWidth: number
+  /** Resolved `overflow-x` on the conversation scroll container. */
+  overflowX: string
+  /** True when the glow's box reaches past the column's content edge — the condition the fix has to survive. */
+  glowBleeds: boolean
+  /**
+   * `scrollWidth - clientWidth`. Deliberately NOT the assertion: `hidden` and
+   * `auto` both report the same value, because `hidden` clips the bleed rather
+   * than reflowing it away. Recorded because it is the vacuity guard in
+   * numbers — it must stay positive at the narrow stops, or the scenario has
+   * stopped reproducing the situation the fix is for.
+   */
+  bleedRange: number
+  /** True when the column still scrolls vertically — the axis the fix must not take away. */
+  scrollsVertically: boolean
+}
+
+/**
+ * Measure the conversation column at the page's current viewport.
+ * @param page - the page under test.
+ * @param width - the viewport width already applied, recorded with the reading.
+ * @returns the stop's overflow relations.
+ */
+function measureColumn(page: Page, width: number): Promise<ColumnMetrics> {
+  return page.evaluate((viewportWidth) => {
+    const scroller = document.querySelector<HTMLElement>('[data-conversation-scroll]')
+    if (scroller === null) throw new Error('conversation scroll container not in the DOM')
+    const glow = scroller.querySelector<SVGElement>('[class*="heroGlow"]')
+    if (glow === null) throw new Error('hero glow not in the DOM — the boot state is not the hero')
+    const box = scroller.getBoundingClientRect()
+    const glowBox = glow.getBoundingClientRect()
+    return {
+      width: viewportWidth,
+      columnWidth: scroller.clientWidth,
+      overflowX: getComputedStyle(scroller).overflowX,
+      // `clientWidth` is the content edge, which is what the scrollable
+      // overflow region is measured against; either side counts as a bleed,
+      // though only the right one can produce a bar in this writing mode.
+      glowBleeds: glowBox.right > box.left + scroller.clientWidth + 0.5 || glowBox.left < box.left - 0.5,
+      bleedRange: scroller.scrollWidth - scroller.clientWidth,
+      scrollsVertically: getComputedStyle(scroller).overflowY === 'auto',
+    }
+  }, width)
+}
+
+/**
+ * Scroll the column sideways the way a user would and report where it landed.
+ *
+ * This is the one signal that separates the two states, and it is why the
+ * scenario needs a real engine: `overflow-x: hidden` leaves the box
+ * programmatically scrollable and leaves `scrollWidth` untouched, so every
+ * property reading agrees across the fix. Only refusing an actual input event
+ * differs — measured at the 1200px stop, the shipped column stays at 0 while
+ * the same page with `overflow-x: auto` forced on lands at its scroll boundary.
+ * @param page - the page under test.
+ * @returns `scrollLeft` after one horizontal wheel over the column.
+ */
+async function wheelHorizontally(page: Page): Promise<number> {
+  const origin = await page.evaluate(() => {
+    const scroller = document.querySelector<HTMLElement>('[data-conversation-scroll]')
+    if (scroller === null) throw new Error('conversation scroll container not in the DOM')
+    // Start from the origin so the reading is this gesture's own effect.
+    scroller.scrollLeft = 0
+    const box = scroller.getBoundingClientRect()
+    // Near the top of the column, clear of the centered hero card: the wheel
+    // must reach the column, not a nested scroller the composer owns.
+    return { x: box.left + box.width / 2, y: box.top + 60 }
+  })
+  await page.mouse.move(origin.x, origin.y)
+  await page.mouse.wheel(WHEEL_DELTA, 0)
+  // A fixed settle, then two frames. Polling for a settled value cannot be
+  // used here — the value under test is 0, which a poll starting at 0 accepts
+  // before the gesture has had any chance to move it — so the wait is
+  // generous enough to cover a smooth-scroll animation on any engine the lane
+  // runs on. The timing is identical on both sides of the mutation control
+  // below, which is what makes a 0 reading evidence rather than a race won.
+  await page.waitForTimeout(400)
+  return page.evaluate(() => new Promise<number>((resolve) => {
+    requestAnimationFrame(() => {
+      requestAnimationFrame(() => {
+        resolve(document.querySelector<HTMLElement>('[data-conversation-scroll]')?.scrollLeft ?? -1)
+      })
+    })
+  }))
+}
+
+/**
+ * Measure the positive horizontal scroll boundary without changing the
+ * shipped overflow mode. This is distinct from `scrollWidth - clientWidth`
+ * when a stable scrollbar gutter leaves part of the overflow on the negative
+ * side of the scroll origin.
+ * @param page - the page under test.
+ * @returns the greatest positive `scrollLeft` reachable by the control gesture.
+ */
+async function horizontalScrollLimit(page: Page): Promise<number> {
+  return page.evaluate((delta) => {
+    const scroller = document.querySelector<HTMLElement>('[data-conversation-scroll]')
+    if (scroller === null) throw new Error('conversation scroll container not in the DOM')
+    const previousScrollBehavior = scroller.style.scrollBehavior
+    scroller.style.scrollBehavior = 'auto'
+    scroller.scrollLeft = delta
+    const limit = scroller.scrollLeft
+    scroller.scrollLeft = 0
+    scroller.style.scrollBehavior = previousScrollBehavior
+    return limit
+  }, WHEEL_DELTA)
+}
+
+/** A stop's readings plus where a horizontal wheel over it landed. */
+type ColumnStop = ColumnMetrics & {
+  /** `scrollLeft` after one horizontal wheel: the user-facing claim, 0 at every stop. */
+  scrollLeftAfterWheel: number
+}
+
+/**
+ * Render the golden body: one line per stop, relations only.
+ *
+ * Absolute pixels are deliberately absent apart from `scrollLeftAfterWheel`,
+ * which the fix pins to 0 by construction. The bleed is recorded as a boolean
+ * rather than its width, so the golden survives any platform whose column
+ * lands a pixel off — a fixture that has to be re-recorded per platform
+ * documents the platform, not the change.
+ * @param stops - the measured stops, in sweep order.
+ * @returns the golden body, without a trailing newline.
+ */
+function renderGeometry(stops: ColumnStop[]): string {
+  return [
+    '# Conversation column horizontal overflow',
+    '',
+    '| viewport | overflow-x | glow bleeds past the column | scrollLeft after a horizontal wheel | scrolls vertically |',
+    '| --- | --- | --- | --- | --- |',
+    ...stops.map(stop => `| ${String(stop.width)}px | ${stop.overflowX} | ${String(stop.glowBleeds)} `
+      + `| ${String(stop.scrollLeftAfterWheel)}px | ${String(stop.scrollsVertically)} |`),
+  ].join('\n')
+}
+
+describe('web e2e: the conversation column scrolls on one axis', () => {
+  let scaffold: WebScaffold
+  let browser: Browser
+  let page: Page
+  let tripwire: ReturnType<typeof watchConsole>
+
+  beforeAll(async () => {
+    scaffold = await launchWebScaffold({})
+    browser = await chromium.launch()
+    page = await newEnglishPage(browser, 900)
+    tripwire = watchConsole(page)
+    await page.goto(scaffold.baseUrl, { waitUntil: 'load' })
+    await page.waitForSelector('[data-conversation-scroll] [class*="heroGlow"]', { timeout: 30_000 })
+  }, 180_000)
+
+  afterAll(async () => {
+    await browser?.close()
+    await scaffold?.close()
+  })
+
+  /**
+   * Resize to a viewport and read the column once its width stops moving.
+   *
+   * The glow rides the hero box, which rides the column, and the frame eases
+   * its column tracks over `--ds-transition-duration-slow`: reading straight
+   * after a resize can report the previous viewport's relation, or a width
+   * caught mid-transition.
+   * @param width - viewport width to settle at.
+   * @returns the column's readings at that width.
+   */
+  const settleAt = async (width: number): Promise<ColumnMetrics> => {
+    await page.setViewportSize({ width, height: 900 })
+    let previous = -1
+    await expect.poll(async () => {
+      const current = (await measureColumn(page, width)).columnWidth
+      const settled = current === previous
+      previous = current
+      return settled
+    }, { timeout: 10_000 }).toBe(true)
+    return measureColumn(page, width)
+  }
+
+  /**
+   * Sweep the stops once per run and hand the SAME readings to every assertion
+   * below, so the golden and the assertions describe one measurement instead of
+   * two runs that could disagree. Memoized rather than re-run per test: the
+   * gestures below move the viewport, and a second sweep would be a second
+   * chance for a resize to settle differently.
+   * @returns the stops in {@link WIDTHS} order.
+   */
+  let swept: Promise<ColumnStop[]> | undefined
+  const sweep = (): Promise<ColumnStop[]> => {
+    swept ??= (async () => {
+      const stops: ColumnStop[] = []
+      for (const width of WIDTHS) {
+        stops.push({ ...await settleAt(width), scrollLeftAfterWheel: await wheelHorizontally(page) })
+      }
+      return stops
+    })()
+    return swept
+  }
+
+  it('never scrolls horizontally, at any width the glow bleeds past', async () => {
+    onTestFailed(() => saveFailureShot(page, 'web-e2e-conversation-column-overflow'))
+    const stops = await sweep()
+    // The vacuity guard, in two halves: the glow has to reach past the column
+    // at the narrow stops, and that reach has to still register as scrollable
+    // overflow. Without both, the claim below holds for free.
+    expect(stops.filter(stop => stop.glowBleeds).map(stop => stop.width)).toEqual([
+      1200, 1000, 800, CONTROL_VIEWPORT,
+    ])
+    for (const stop of stops.filter(stop => stop.glowBleeds)) {
+      expect(stop.bleedRange, `viewport ${String(stop.width)}`).toBeGreaterThan(0)
+    }
+    for (const stop of stops) {
+      expect(stop.overflowX, `viewport ${String(stop.width)}`).toBe('hidden')
+      // The reported symptom, stated directly: a horizontal wheel over the
+      // column moves nothing, at every stop.
+      expect(stop.scrollLeftAfterWheel, `viewport ${String(stop.width)}`).toBe(0)
+      // The axis the column is a scroller for must survive the fix.
+      expect(stop.scrollsVertically, `viewport ${String(stop.width)}`).toBe(true)
+    }
+    expect(tripwire.pageErrors).toEqual([])
+  }, 120_000)
+
+  it('reports the pre-fix state when the axis is opened back up', async () => {
+    onTestFailed(() => saveFailureShot(page, 'web-e2e-conversation-column-overflow-control'))
+    // The mutation control, run in the page rather than against a second
+    // build: it restores exactly what the fix changed — the initial `visible`
+    // that a one-axis scroller computes to `auto` — and shows the same gesture,
+    // at the same timing, carrying the column to its positive scroll boundary.
+    // Without it a `scrollLeft` of 0 could equally mean the wheel never arrived.
+    // Injected with an id rather than through `addStyleTag`, so the teardown
+    // below can take the sheet out again by selector: it must not outlive this
+    // test, or the golden ends up reading the control.
+    await page.evaluate((id: string) => {
+      const sheet = document.createElement('style')
+      sheet.id = id
+      sheet.textContent = '[data-conversation-scroll] { overflow-x: auto !important; }'
+      document.head.append(sheet)
+    }, CONTROL_STYLE_ID)
+    try {
+      // Resolve the mutated layout at the narrowest sweep stop. At wider stops,
+      // a classic scrollbar can change the available box enough to remove the
+      // overflow that the control is meant to expose.
+      const before = await settleAt(CONTROL_VIEWPORT)
+      expect(before.overflowX).toBe('auto')
+      expect(before.bleedRange).toBeGreaterThan(0)
+      const scrollLimit = await horizontalScrollLimit(page)
+      // The control has a reachable horizontal range, and the gesture exceeds
+      // it so the equality below proves that the wheel reached the far edge.
+      expect(scrollLimit).toBeGreaterThan(0)
+      expect(scrollLimit).toBeLessThan(WHEEL_DELTA)
+      // Rounded: `scrollLeft` is fractional under a fractional layout while
+      // the claim is that the column reached the positive boundary, not that
+      // two engines agree on a sub-pixel.
+      expect(Math.round(await wheelHorizontally(page))).toBe(Math.round(scrollLimit))
+    } finally {
+      await page.evaluate((id: string) => {
+        document.getElementById(id)?.remove()
+      }, CONTROL_STYLE_ID)
+    }
+    // The override is gone and the shipped state is back: the later goldens
+    // read the product, not the control.
+    expect((await settleAt(CONTROL_VIEWPORT)).overflowX).toBe('hidden')
+    expect(tripwire.pageErrors).toEqual([])
+  }, 120_000)
+
+  it('matches the committed column-overflow golden', async () => {
+    onTestFailed(() => saveFailureShot(page, 'web-e2e-conversation-column-overflow-golden'))
+    await compareOrRefreshGolden(GEOMETRY_EXPECTED, renderGeometry(await sweep()), MODE)
+    expect(tripwire.pageErrors).toEqual([])
+  }, 120_000)
+
+  it('commits exactly the fixtures it reads', async () => {
+    // No model calls, so no replay log: the golden is the whole inventory.
+    await assertFixtureInventory(SNAPSHOT_DIR, ['geometry.expected.md'])
+  })
+
+  it.skipIf(MODE === 'record')('issued zero model calls and stayed clean', () => {
+    expect(tripwire.warnings).toEqual([])
+    expect(tripwire.pageErrors).toEqual([])
+  })
+})

apps/web/tests/core-web-profile.snapshot.ts

@@ -3,21 +3,27 @@ import { join } from 'node:path'
 import { fileURLToPath } from 'node:url'
 import { afterAll, beforeAll, describe, expect, it } from 'vitest'
 import type { AgentHandle } from '@deepseek-ai/dsh-agent'
-import { CallId } from '@deepseek-ai/dsh-llm'
+import { CallId, createUserMessage } from '@deepseek-ai/dsh-llm'
 import { SessionId } from '@deepseek-ai/dsh-session'
-import { launchWebScaffold, type WebScaffold } from './scaffold.ts'
+import { assertFixtureInventory, launchWebScaffold, type WebScaffold } from './scaffold.ts'
 
 const CORE_WEB_OVERLAY = fileURLToPath(new URL('../../cli/config/core-web.cordis.yml', import.meta.url))
+const SNAPSHOT_DIR = fileURLToPath(new URL('./snapshots/core-web-profile', import.meta.url))
+const FIXTURE = join(SNAPSHOT_DIR, 'session.jsonl')
+const PROMPT = 'Reply exactly CORE_WEB_REQUEST_OK and stop.'
 
 describe('core Web profile', () => {
   let scaffold: WebScaffold
   let agentHandle: AgentHandle
 
   beforeAll(async () => {
-    scaffold = await launchWebScaffold({
-      extraOverlayPath: CORE_WEB_OVERLAY,
-      toolsMode: 'native',
-    })
+    const systemPrompt = process.env.DSH_SYSTEM_PROMPT
+    Reflect.deleteProperty(process.env, 'DSH_SYSTEM_PROMPT')
+    try {
+      scaffold = await launchWebScaffold({ extraOverlayPath: CORE_WEB_OVERLAY, replayFixture: FIXTURE })
+    } finally {
+      if (systemPrompt !== undefined) process.env.DSH_SYSTEM_PROMPT = systemPrompt
+    }
     agentHandle = await scaffold.ctx.agents.create({
       sessionId: SessionId('core-web-profile-smoke'),
       meta: { cwd: scaffold.workspaceCwd },
@@ -33,7 +39,16 @@ describe('core Web profile', () => {
     if (failures.length > 1) throw new AggregateError(failures, 'core Web profile smoke teardown failed')
   })
 
-  it('boots and executes both tools through the shipped Web composition', async () => {
+  it('sends the RL prompt and tool schemas through a real request, then executes both tools', async () => {
+    agentHandle.agent.followup(createUserMessage({
+      content: [{ type: 'text', text: PROMPT }],
+      source: { kind: 'user' },
+    }))
+    await agentHandle.agent.whenIdle()
+
+    const requestHeader = agentHandle.agent.session.requestHeader()
+    if (requestHeader === undefined) throw new Error('the core Web agent issued no model request')
+
     const seedPath = join(scaffold.workspaceCwd, 'profile-smoke.txt')
     await writeFile(seedPath, 'CORE_WEB_EDITOR_OK\n')
     const signal = new AbortController().signal
@@ -60,7 +75,8 @@ describe('core Web profile', () => {
       .trimEnd()
 
     expect({
-      tools: scaffold.ctx.tools.schemas().map(tool => tool.name),
+      prompt: requestHeader.system,
+      tools: requestHeader.tools?.map(tool => tool.name),
       bash: text(bash),
       editor: text(editor),
     }).toMatchInlineSnapshot(`
@@ -69,16 +85,53 @@ describe('core Web profile', () => {
         "editor": "Here's the content of {{cwd}}/profile-smoke.txt with line numbers (which has a total of 2 lines):
            1  CORE_WEB_EDITOR_OK
            2",
+        "prompt": "You are a helpful software engineer assistant.",
         "tools": [
           "bash",
           "str_replace_editor",
         ],
       }
     `)
+    expect(requestHeader.tools).toEqual(scaffold.ctx.tools.schemas(agentHandle.agent))
 
     const entries = [...scaffold.ctx.loader.entries()]
     expect(entries.find(entry => entry.options.id === 'persistent-bash')?.fiber).toBeDefined()
     expect(entries.find(entry => entry.options.id === 'pty-local')?.fiber).toBeDefined()
     expect(entries.find(entry => entry.options.id === 'str-replace-editor')?.fiber).toBeDefined()
+    expect(entries.find(entry => entry.options.id === 'web-runtime')?.fiber).toBeDefined()
+    expect(entries.find(entry => entry.options.id === 'workspace-context')?.fiber).toBeUndefined()
+    await assertFixtureInventory(SNAPSHOT_DIR, ['session.jsonl'])
+  })
+
+  it('uses DSH_SYSTEM_PROMPT as the complete prompt when configured', async () => {
+    const previous = process.env.DSH_SYSTEM_PROMPT
+    process.env.DSH_SYSTEM_PROMPT = 'RL prompt override'
+    let overrideScaffold: WebScaffold | undefined
+    let overrideAgent: AgentHandle | undefined
+    try {
+      overrideScaffold = await launchWebScaffold({ extraOverlayPath: CORE_WEB_OVERLAY, replayFixture: FIXTURE })
+      overrideAgent = await overrideScaffold.ctx.agents.create({
+        sessionId: SessionId('core-web-profile-override'),
+        meta: { cwd: overrideScaffold.workspaceCwd },
+        agentOptions: { provider: 'deepseek-official', model: 'deepseek-v4-flash' },
+      })
+      overrideAgent.agent.followup(createUserMessage({
+        content: [{ type: 'text', text: PROMPT }],
+        source: { kind: 'user' },
+      }))
+      await overrideAgent.agent.whenIdle()
+      expect(overrideAgent.agent.session.requestHeader()?.system).toBe('RL prompt override')
+    } finally {
+      try {
+        await overrideAgent?.dispose()
+      } finally {
+        try {
+          await overrideScaffold?.close()
+        } finally {
+          if (previous === undefined) Reflect.deleteProperty(process.env, 'DSH_SYSTEM_PROMPT')
+          else process.env.DSH_SYSTEM_PROMPT = previous
+        }
+      }
+    }
   })
 })

apps/web/tests/onboarding-deepseek-config.e2e.ts

@@ -161,6 +161,58 @@ describe.skipIf(MODE === 'record')('web e2e: first-run DeepSeek credential setup
     expect(tripwire.pageErrors).toEqual([])
   }, 60_000)
 
+  it('never paints the takeover chrome on a configured reload, even with the settings join held open', async () => {
+    onTestFailed(() => saveFailureShot(page, 'web-e2e-onboarding-configured-reload'))
+    // Regression pin for the reload white flash: both steps are satisfied
+    // (welcome acknowledged, credential configured), yet each must LOAD its
+    // private join before it can decide not to show. The chrome lives inside
+    // the step (OnboardingSurface), so the deciding window paints and blocks
+    // nothing. Holding settings.describe widens that window from loopback
+    // RTT scale to a deterministic hundreds of milliseconds, removing all
+    // timing dependence from the sampler assertions below.
+    //
+    // The sampler init script persists across this shared page's later
+    // navigations (init scripts re-run per navigation); that stays harmless
+    // because no later scenario in this file legitimately shows the
+    // takeover, and only this test reads __takeoverSightings.
+    await page.addInitScript(() => {
+      const sightings: string[] = []
+      ;(window as unknown as { __takeoverSightings: string[] }).__takeoverSightings = sightings
+      setInterval(() => {
+        if (document.querySelector('[class*="onboardingStage"], [class*="onboardingMask"]') !== null) {
+          sightings.push('chrome')
+        }
+        if (document.getElementById('root')?.inert === true) sightings.push('inert')
+      }, 8)
+    })
+    // EVERY settings.describe issued before the release is held — not just
+    // the first — so the pin cannot silently collapse back to loopback
+    // timing if a second boot-time consumer of the join ever appears.
+    let released = false
+    const heldRoutes: Array<() => void> = []
+    const releaseDescribe = (): void => {
+      released = true
+      for (const resolve of heldRoutes.splice(0)) resolve()
+    }
+    await page.route('**/api/settings.describe', async (route) => {
+      if (!released) await new Promise<void>((resolve) => { heldRoutes.push(resolve) })
+      await route.continue()
+    })
+    const warningsBefore = tripwire.warnings.length
+    await page.reload({ waitUntil: 'commit' })
+    await page.waitForSelector('[class*="frame"]', { timeout: 15_000 })
+    // The app is painted and interactive while the steps are still deciding.
+    await page.waitForTimeout(600)
+    releaseDescribe()
+    await page.waitForTimeout(400)
+    await page.unroute('**/api/settings.describe')
+    acknowledgeReloadConnectionLoss(tripwire, warningsBefore)
+    expect(await page.evaluate(() =>
+      (window as unknown as { __takeoverSightings: string[] }).__takeoverSightings)).toEqual([])
+    expect(await page.locator('[class*="onboardingStage"]').count()).toBe(0)
+    expect(tripwire.pageErrors).toEqual([])
+  }, 60_000)
+
   it('configures arbitrary DeepSeek models and prompts after the selected model is removed', async () => {
     onTestFailed(() => saveFailureShot(page, 'web-e2e-onboarding-deepseek-models'))
     // Opened here rather than inherited: the credential test reloads the page

apps/web/tests/scaffold.ts

@@ -33,7 +33,13 @@ import { Context } from 'cordis'
 import Loader from '@cordisjs/plugin-loader'
 import Include, { type PatchOptions } from '@cordisjs/plugin-include'
 import { scrubRequestHeaders } from '@deepseek-ai/dsh-acp-snapshot'
-import { assertEntriesLoaded, loadOverlayPatches } from '@deepseek-ai/dsh-app-boot'
+import {
+  addHarnessSourceSection,
+  assertEntriesLoaded,
+  composeEntries,
+  healProfilesModuleFallback,
+  loadOverlayPatches,
+} from '@deepseek-ai/dsh-app-boot'
 import { dshHomePath } from '@deepseek-ai/dsh-paths'
 import {
   WELCOME_NOTICE_ACK_FIELD, WELCOME_NOTICE_SETTINGS_NAMESPACE, WELCOME_NOTICE_VERSION,
@@ -54,7 +60,6 @@ import * as ToolCordis from '@deepseek-ai/dsh-tool-cordis'
 // Empty type imports carry the httpServer/agents/sessionPersistence Context merges.
 import type {} from '@deepseek-ai/dsh-host-webserver'
 import type {} from '@deepseek-ai/dsh-agent'
-import { addHarnessSourceSection, healProfilesModuleFallback } from '@deepseek-ai/dsh-app-boot'
 import { REPO_ROOT, requireDist } from './support.ts'
 
 /** Snapshot mode for the lane, from $DSH_SNAPSHOT (same vocabulary as the other snapshot suites). */
@@ -252,6 +257,11 @@ export async function launchWebScaffold(options: LaunchOptions = {}): Promise<We
   const extraOverlayPatches = options.extraOverlayPath === undefined
     ? []
     : loadOverlayPatches('web e2e scaffold', options.extraOverlayPath)
+  const composedRows = composeEntries([basePatches, surfacePatches, extraOverlayPatches])
+  const webRuntimeConfig = composedRows.find(row => row.id === 'web-runtime')?.config as {
+    surfaceContext?: boolean
+  } | undefined
+  const surfaceContext = webRuntimeConfig?.surfaceContext !== false
   const patches: PatchOptions[] = [
     ...basePatches,
     ...surfacePatches,
@@ -290,7 +300,9 @@ export async function launchWebScaffold(options: LaunchOptions = {}): Promise<We
     },
     // The bundle's web-runtime row resolves the same built dist under test
     // (apps/web IS @deepseek-ai/dsh-frontend); only the URL line is silenced.
-    { id: 'web-runtime', config: { mode: 'production', printUrl: false } },
+    // Preserve the composed surface-context choice because a patch replaces
+    // the row's complete config.
+    { id: 'web-runtime', config: { mode: 'production', printUrl: false, surfaceContext } },
     ...options.remoteAuthority === undefined
       ? []
       : [{ id: 'connection', config: { trustedHosts: [options.remoteAuthority] } }],
@@ -346,7 +358,9 @@ export async function launchWebScaffold(options: LaunchOptions = {}): Promise<We
     // The shipped CLI deliberately has no dependency on this opt-in package.
     // Keep the Loader row real without broadening the product installation.
     if (options.cordisTools === true) ctx.loader.builtins['tool-cordis'] = ToolCordis
-    ctx.inject(['systemPrompt'], (promptCtx) => { addHarnessSourceSection(promptCtx, REPO_ROOT) })
+    if (surfaceContext) {
+      ctx.inject(['systemPrompt'], (promptCtx) => { addHarnessSourceSection(promptCtx, REPO_ROOT) })
+    }
     await ctx.loader.create({
       name: 'cordis:include',
       config: { path: pathToFileURL(rootConfig).href, patches },

apps/web/tests/snapshots/composer-tab-geometry/geometry.expected.md

@@ -2,7 +2,7 @@
 
 ## Wide viewport (1680px, card at its cap)
 
-- Chat: scrollbar-gutter stable, overflow auto/auto
+- Chat: scrollbar-gutter stable, overflow hidden/auto
 - Chat scroller scrolls: true
 - Chat reserved band: 8px
 - Trajectory: scrollbar-gutter stable, overflow hidden/auto
@@ -14,7 +14,7 @@
 
 ## Narrow viewport (800px, card shrinking with the column)
 
-- Chat: scrollbar-gutter stable, overflow auto/auto
+- Chat: scrollbar-gutter stable, overflow hidden/auto
 - Chat scroller scrolls: true
 - Chat reserved band: 8px
 - Trajectory: scrollbar-gutter stable, overflow hidden/auto
@@ -26,7 +26,7 @@
 
 ## Wide viewport, reservation removed in the page (control)
 
-- Chat: scrollbar-gutter auto, overflow auto/auto
+- Chat: scrollbar-gutter auto, overflow hidden/auto
 - Chat scroller scrolls: true
 - Chat reserved band: 8px
 - Trajectory: scrollbar-gutter auto, overflow hidden/hidden

apps/web/tests/snapshots/conversation-column-overflow/geometry.expected.md

@@ -0,0 +1,9 @@
+# Conversation column horizontal overflow
+
+| viewport | overflow-x | glow bleeds past the column | scrollLeft after a horizontal wheel | scrolls vertically |
+| --- | --- | --- | --- | --- |
+| 1680px | hidden | false | 0px | true |
+| 1200px | hidden | true | 0px | true |
+| 1000px | hidden | true | 0px | true |
+| 800px | hidden | true | 0px | true |
+| 600px | hidden | true | 0px | true |

apps/web/tests/snapshots/core-web-profile/session.jsonl

@@ -0,0 +1,7 @@
+{"type":"session","version":0,"id":"{{sessionId}}","createdAt":1785974400000,"cwd":"{{cwd}}"}
+{"type":"user/message","seq":0,"time":1785974400001,"data":{"content":[{"type":"text","text":"Reply exactly CORE_WEB_REQUEST_OK and stop."}],"source":{"kind":"user","rpcId":"{{rpcId}}"}},"surfaceOp":"append"}
+{"type":"assistant/chunk","seq":1,"time":1785974400002,"data":{"turn":1,"step":1,"chunk":{"type":"block-start","index":0,"blockType":"text"}}}
+{"type":"assistant/chunk","seq":2,"time":1785974400003,"data":{"turn":1,"step":1,"chunk":{"type":"text-delta","index":0,"text":"CORE_WEB_REQUEST_OK"}}}
+{"type":"assistant/chunk","seq":3,"time":1785974400004,"data":{"turn":1,"step":1,"chunk":{"type":"block-end","index":0,"block":{"type":"text","text":"CORE_WEB_REQUEST_OK"}}}}
+{"type":"assistant/chunk","seq":4,"time":1785974400005,"data":{"turn":1,"step":1,"chunk":{"type":"usage","usage":{"inputTokens":10,"outputTokens":4}}}}
+{"type":"assistant/chunk","seq":5,"time":1785974400006,"data":{"turn":1,"step":1,"chunk":{"type":"finish","reason":{"kind":"stop"}}}}

apps/web/tsconfig.json

@@ -44,6 +44,7 @@
     "tests/hmr-live.e2e.ts",
     "tests/seeded-history.e2e.ts",
     "tests/sidebar-scrollbar.e2e.ts",
+    "tests/conversation-column-overflow.e2e.ts",
     "tests/code-mode-round.e2e.ts",
     "tests/composer-draft-scroll.e2e.ts",
     "tests/cordis-tool-round.e2e.ts",