@xingular/harness 1.5.0-unstable → 1.6.0-unstable

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (95) hide show
  1. package/CHANGELOG.md +93 -0
  2. package/browser-extension/README.md +32 -0
  3. package/browser-extension/background.js +612 -0
  4. package/browser-extension/manifest.json +16 -0
  5. package/browser-extension/overlay.js +97 -0
  6. package/browser-extension/popup.css +19 -0
  7. package/browser-extension/popup.html +30 -0
  8. package/browser-extension/popup.js +101 -0
  9. package/browser-extension/site-access.js +63 -0
  10. package/dist/cli.js +1489 -680
  11. package/dist/sdk/browser/config.d.ts +3 -0
  12. package/dist/sdk/browser/config.js +2 -1
  13. package/dist/sdk/browser/extension/bridge.d.ts +55 -0
  14. package/dist/sdk/browser/extension/bridge.js +336 -0
  15. package/dist/sdk/browser/extension/session.d.ts +50 -0
  16. package/dist/sdk/browser/extension/session.js +321 -0
  17. package/dist/sdk/browser/index.d.ts +7 -2
  18. package/dist/sdk/browser/index.js +58 -25
  19. package/dist/sdk/browser/serverMethods.js +4 -1
  20. package/dist/sdk/browser/tools.d.ts +99 -2
  21. package/dist/sdk/browser/tools.js +6 -1
  22. package/dist/sdk/commands/registry.d.ts +5 -0
  23. package/dist/sdk/commands/registry.js +1 -0
  24. package/dist/sdk/harness.d.ts +4 -0
  25. package/dist/sdk/harness.js +1 -1
  26. package/dist/sdk/pi/retryPolicy.d.ts +19 -0
  27. package/dist/sdk/pi/retryPolicy.js +11 -0
  28. package/dist/sdk/prompt/compile.d.ts +0 -2
  29. package/dist/sdk/prompt/compile.js +2 -6
  30. package/dist/sdk/prompt/reporting.d.ts +1 -1
  31. package/dist/sdk/prompt/reporting.js +30 -13
  32. package/dist/sdk/prompt/standard.d.ts +1 -1
  33. package/dist/sdk/prompt/standard.js +1 -1
  34. package/dist/sdk/remote/webClient.js +6 -2
  35. package/dist/sdk/sessions/contracts.d.ts +2 -0
  36. package/dist/sdk/sessions/contracts.js +5 -1
  37. package/dist/sdk/sessions/planMode.d.ts +16 -6
  38. package/dist/sdk/sessions/planMode.js +59 -23
  39. package/dist/sdk/sessions/runner.js +14 -5
  40. package/dist/sdk/sessions/workSlice.d.ts +15 -1
  41. package/dist/sdk/sessions/workSlice.js +17 -9
  42. package/dist/sdk/terminal/controller.d.ts +6 -0
  43. package/dist/sdk/terminal/controller.js +44 -5
  44. package/dist/sdk/terminal/opentui/model.js +3 -1
  45. package/dist/sdk/terminal/piExtension.js +7 -2
  46. package/dist/sdk/terminal/stallWatch.d.ts +18 -0
  47. package/dist/sdk/terminal/stallWatch.js +20 -0
  48. package/dist/sdk/terminal/subagentMonitor.d.ts +2 -0
  49. package/dist/sdk/terminal/subagentMonitor.js +4 -1
  50. package/dist/sdk/terminal/terminals.js +35 -4
  51. package/dist/sdk/terminal/toolName.d.ts +6 -0
  52. package/dist/sdk/terminal/toolName.js +12 -0
  53. package/dist/sdk/terminal/tui/commandPages.d.ts +1 -1
  54. package/dist/sdk/terminal/tui/commandPages.js +3 -0
  55. package/dist/sdk/terminal/tui/commands.d.ts +14 -0
  56. package/dist/sdk/terminal/tui/commands.js +4 -1
  57. package/dist/sdk/terminal/tui/pages/followUpQueuePage.d.ts +34 -0
  58. package/dist/sdk/terminal/tui/pages/followUpQueuePage.js +130 -0
  59. package/dist/sdk/terminal/tui/pages/trajectoryPage.js +37 -7
  60. package/dist/sdk/terminal/tui/runtime.js +2 -1
  61. package/dist/sdk/terminal/tui/settingsModel.js +88 -7
  62. package/dist/sdk/tools/askUser.js +2 -2
  63. package/dist/sdk/world/world.d.ts +12 -2
  64. package/dist/sdk/world/world.js +42 -4
  65. package/dist/terminal-opentui.js +1456 -756
  66. package/dist/terminal-pi.js +62 -31
  67. package/docs/README.md +1 -0
  68. package/docs/architecture/02-sessions-and-runs.md +43 -25
  69. package/docs/architecture/04-prompt.md +23 -22
  70. package/docs/architecture/10-server-protocol.md +23 -3
  71. package/docs/architecture/11-terminal.md +41 -29
  72. package/docs/architecture/12-storage.md +2 -1
  73. package/docs/architecture/13-world-tools.md +35 -22
  74. package/docs/architecture/14-surfaces-and-parity.md +24 -20
  75. package/docs/architecture/README.md +4 -1
  76. package/docs/architecture/proposals/PROP-0026-connected-sketch-display.md +3 -1
  77. package/docs/decisions/ADR-0027-connected-sketch-display.md +4 -2
  78. package/docs/decisions/ADR-0076-codex-final-answer-verbatim.md +6 -1
  79. package/docs/decisions/ADR-0077-plan-runs-may-conclude-no-plan-needed.md +120 -0
  80. package/docs/decisions/ADR-0078-child-aware-stall-watch.md +46 -0
  81. package/docs/decisions/ADR-0079-phase-led-reporting-guidance.md +72 -0
  82. package/docs/decisions/ADR-0080-fixed-width-question-sketch-canvas.md +50 -0
  83. package/docs/decisions/ADR-0081-no-work-slice-budget-by-default.md +36 -0
  84. package/docs/decisions/ADR-0082-connected-brave-browser.md +72 -0
  85. package/docs/parity.md +7 -5
  86. package/docs/releases/1.5.0-unstable.md +12 -3
  87. package/docs/releases/1.5.1-unstable.md +24 -0
  88. package/docs/releases/1.6.0-unstable.md +59 -0
  89. package/docs/study/00-index.md +17 -1
  90. package/docs/study/STUDY-0007-claude-code-prompt.md +162 -0
  91. package/docs/study/STUDY-0008-browser-control.md +281 -0
  92. package/docs/study/STUDY-0009-claude-code-reporting.md +205 -0
  93. package/docs/study/sources/claude-code.md +51 -39
  94. package/docs/terminal-tui.md +46 -18
  95. package/package.json +2 -1
@@ -0,0 +1,72 @@
1
+ ---
2
+ id: ADR-0079
3
+ title: Phase-led flexible reporting guidance
4
+ status: accepted
5
+ created: 2026-11-02
6
+ updated: 2026-11-02
7
+ amends: ADR-0076 (report phases, reader context, verification honesty, subagent reports and headings)
8
+ related:
9
+ - ADR-0076
10
+ - STUDY-0009
11
+ ---
12
+ # ADR-0079 — Phase-led flexible reporting guidance
13
+
14
+ > Amends [ADR-0076](ADR-0076-codex-final-answer-verbatim.md): keep its self-contained, outcome-first
15
+ > report and flexible structure, while strengthening the reader, evidence, heading and subagent
16
+ > guidance based on [STUDY-0009](../study/STUDY-0009-claude-code-reporting.md).
17
+
18
+ ## Context
19
+
20
+ The study compared Claude Code's external and internal prompt phases with the Harness reporting
21
+ section. It found three useful gaps: the Harness did not explicitly help a reader who has lost the
22
+ thread, did not say directly how to report checks that were not run or failed, and gave subagents no
23
+ caller-facing report contract. The in-flight progress sentence also appeared in both `writing` and
24
+ `reporting`, although `reporting` shapes the final answer.
25
+
26
+ The user asked for phase-led final-report guidance and frequent Markdown `#` / `##` headings such as
27
+ "What changed", "Checks" and "Things to know", while explicitly rejecting a fixed heading vocabulary.
28
+
29
+ ## Decision
30
+
31
+ - Keep the final reply self-contained, outcome-first, evidence-led and scaled to the work. Organize
32
+ the guidance in final-report order: reader and purpose, opening, body, evidence and honesty,
33
+ person-owned next step when one exists, headings, lists and tables, length, references and
34
+ formatting, then examples and pre-send review. Follow citation and file-link rules already
35
+ supplied by the active session or host. Pre-tool narration and in-flight updates remain in their existing prompt section;
36
+ remove only the duplicate progress-update sentence from `reporting`. Do not add an away-recap
37
+ surface or numeric word caps.
38
+ - Add the cold-reader test: assume the person may have stepped away and lost the thread; give enough
39
+ context to pick the work back up and match the explanation to cues about their experience.
40
+ - State verification honestly. Say when a check was not run, never imply that it passed, report
41
+ failures with relevant output, and state passes plainly.
42
+ - When writing as a subagent, give the caller a concise account of work done and key findings. The
43
+ caller integrates and relays it. Include absolute paths and useful evidence, snippets only when
44
+ exact text is load-bearing, and no recap of code merely read.
45
+ - For substantive replies with distinct parts, often use Markdown headings: a `#` title when useful
46
+ and `##` headings for meaningful sections. Choose labels from the reply's actual content. Examples
47
+ such as "What changed", "Checks" and "Things to know" are optional wording, not a fixed list or
48
+ template. Short replies may stay in prose; headings are never forced and empty sections are not
49
+ added.
50
+ - Revise one worked example to demonstrate a content-specific `#` / `##` structure and retain a
51
+ short heading-free example. Keep the 12,000-character reporting budget.
52
+
53
+ ## Consequences
54
+
55
+ - Reporting behavior changes the stable prompt prefix. Existing live terminal sessions retain the
56
+ pack they started with; new sessions receive the updated text. Cache reuse after the prefix change
57
+ must be measured only with real application requests and the account owner's permission
58
+ (INV-PRM-8).
59
+ - No storage, server protocol or tool capability changes.
60
+ - The guidance favors headings for substantial multi-part reports without prescribing section names
61
+ or forcing headings on simple replies.
62
+
63
+ ## Alternatives
64
+
65
+ - **Require a fixed set of report headings.** Rejected: the user wants headings chosen for the
66
+ content, and fixed labels become a template.
67
+ - **Keep reports prose-only unless a heading is requested.** Rejected: substantial reports should
68
+ use headings often enough to make their parts easy to find.
69
+ - **Move progress updates or add an away recap to reporting.** Rejected: in-flight updates already
70
+ belong in `writing`, and this harness has no away-recap surface.
71
+ - **Adopt Claude Code's numeric word limits.** Rejected: the study did not establish their quality
72
+ effect for this Harness.
@@ -0,0 +1,50 @@
1
+ ---
2
+ id: ADR-0080
3
+ title: Use a fixed 80-column canvas for question sketches
4
+ status: accepted
5
+ created: 2026-11-28
6
+ updated: 2026-11-28
7
+ related: [PROP-0026, ADR-0027, ADR-0029, INV-PAR-1, INV-PAR-2]
8
+ supersedes: null
9
+ superseded_by: null
10
+ ---
11
+
12
+ # ADR-0080 — Use a fixed 80-column canvas for question sketches
13
+
14
+ ## Context
15
+
16
+ The shared question-sketch contract permits up to 80 printable ASCII columns, but neither validation
17
+ nor the tool guidance required the model to use that full width. The terminal's display width also
18
+ followed the viewport, so the same sketch could appear at different widths on different terminals.
19
+ A fixed design canvas gives the model room for contextual UI detail and makes terminal previews
20
+ consistent without breaking callers that provide shorter sketches.
21
+
22
+ ## Decision
23
+
24
+ Ask the model to design detailed `uiSketch` wireframes on a fixed 80-column canvas. This remains
25
+ guidance: preserve the existing validator, which allows sketches shorter than 80 columns and does
26
+ not require trailing spaces.
27
+
28
+ The terminal projects every valid source sketch to 80 content columns, padding shorter rows and
29
+ extending standalone divider rows to the same width. Keep the outer preview frame no wider than the
30
+ available dock; the content remains horizontally pannable when the canvas does not fit. Keep the
31
+ existing structural-border projection within the canvas. Padding and projected glyphs are display-only;
32
+ source events and answer payloads remain unchanged.
33
+
34
+ The App may keep its literal ASCII preview and overflow behavior. Both surfaces use the same source
35
+ contract and model guidance; identical preview typography is not required. This amends only the
36
+ viewport-relative canvas and divider sizing in [ADR-0027](ADR-0027-connected-sketch-display.md) /
37
+ [PROP-0026](../architecture/proposals/PROP-0026-connected-sketch-display.md). Their structural
38
+ recognition and persistence rules remain in force.
39
+
40
+ ## Consequences
41
+
42
+ Short source rows appear padded in the terminal but remain unchanged in history. A model may still
43
+ return a narrower drawing because canvas usage is not schema-enforced. There is no data migration,
44
+ wire-contract change, or answer-payload change.
45
+
46
+ ## Verification
47
+
48
+ Cover fixed width and padding in unit tests; verify the viewport-bounded frame and horizontal panning
49
+ at compact, standard, and wide PTY sizes. Confirm persisted sketch text and submitted answers remain
50
+ unchanged. Fixture verification does not guarantee every live-model sketch uses the full canvas.
@@ -0,0 +1,36 @@
1
+ ---
2
+ id: ADR-0081
3
+ title: No work-slice budget by default
4
+ status: accepted
5
+ created: 2026-11-02
6
+ updated: 2026-11-02
7
+ related: [PROP-0041, PROP-0044, ADR-0070, ADR-0075, INV-RUN-13]
8
+ supersedes: null
9
+ superseded_by: null
10
+ ---
11
+
12
+ # ADR-0081 — No work-slice budget by default
13
+
14
+ ## Context
15
+
16
+ PROP-0041 bounded a reply at a number of tool calls and minutes, then handed off, and ADR-0070 and
17
+ PROP-0044 tuned that to 320 calls and 60 minutes with up to eight continued slices. In practice a
18
+ reply that was plainly still working was cut mid-task, told to "stop using tools and write a short
19
+ report", and showed as unfinished while a child worker kept the run alive.
20
+
21
+ ## Decision
22
+
23
+ The foundation default has no tool-call, time or token limit. `DEFAULT_SLICE_POLICY` sets
24
+ `maxToolCalls`, `maxMinutes` and `maxNewTokens` to 0, so `sliceLimitReached` never fires and no reply
25
+ ends or continues at a budget. The guard, `workSlice` on `sessions.run` and `XINGULAR_WORK_SLICE`
26
+ stay for a caller that wants a limit, and the three-compaction guard is unchanged. The model is no
27
+ longer told a budget exists, and the standard prompt asks it to keep working until the request is
28
+ done or the person interrupts.
29
+
30
+ ## Consequences
31
+
32
+ - The stable prompt prefix changes once, so the first reply after upgrading is uncached.
33
+ - Runaway work is bounded only by the stall watch (ADR-0054), the compaction guard, approvals and
34
+ the person's Esc; a host that wants a ceiling sets `workSlice`.
35
+ - The budget code, its protocol field and its heartbeat fields remain, so a later decision can
36
+ delete them without a migration. `heartbeat.maxToolCalls` is 0.
@@ -0,0 +1,72 @@
1
+ ---
2
+ id: ADR-0082
3
+ title: Connected Brave browser backend
4
+ status: accepted
5
+ created: 2026-11-20
6
+ updated: 2026-11-20
7
+ related: [STUDY-0008, PROP-0040, INV-PAR-1, INV-PAR-2]
8
+ supersedes: null
9
+ superseded_by: null
10
+ ---
11
+
12
+ # ADR-0082 — Connected Brave browser backend
13
+
14
+ ## Context
15
+
16
+ The isolated Chromium browser cannot use the person's existing sign-in. The App's WebKit tabs are
17
+ host-owned, have a different profile lifecycle, and do not implement snapshot-scoped ref actions.
18
+ The person asked to choose between an isolated browser and their existing signed-in Brave profile,
19
+ with no automatic switch if the chosen backend disconnects.
20
+
21
+ ## Decision
22
+
23
+ Add an opt-in Brave backend through an unpacked Manifest V3 extension. The extension uses
24
+ `chrome.debugger` to adopt existing normal tabs, and the Harness adapts the extension to the shared
25
+ `BrowserToolSession` and `browser_*` schemas. The extension is loaded manually and pairs with the
26
+ Harness over an authenticated loopback bridge. Do not enable remote debugging on the person's
27
+ profile and do not require Native Messaging. Pairing validates the extension origin; subsequent
28
+ service-worker calls require the bearer token because the browser may omit Origin on those requests,
29
+ and any supplied mismatched Origin is rejected.
30
+
31
+ Browser backend selection is device-local. A project, variant or page cannot enable Brave. A person
32
+ must explicitly choose Brave, pair the extension with a short-lived code, and allow each exact
33
+ origin from the extension popup. Chrome host grants can cover every port on a host, so the extension
34
+ also records the exact scheme, host and port and checks that allow list before exposing a tab.
35
+ Incognito is disabled. The extension does
36
+ not request history or cookie permissions and does not export cookies; a signed-in page remains
37
+ signed in because it uses the existing profile. The extension's `debugger` permission is broad and
38
+ must be disclosed before the person selects Brave.
39
+
40
+ While connected and permitted, the page shows **“Xingular Agent is on this site”**, with Pause and
41
+ Stop controls and a brief pulse at the most recent click. Pause refuses new calls but lets an action
42
+ already in flight finish. Stop revokes the pairing, rejects pending calls, clears the badge and closes
43
+ the bridge. Normal Harness shutdown closes the bridge but keeps the pairing so the extension can
44
+ reconnect. A lost or paused connection refuses calls; it never selects another browser. A timed-out
45
+ call reports that its outcome is unknown and requires a fresh snapshot before retry.
46
+
47
+ The shared browser gate remains authoritative: snapshot refs are invalidated when page state or the
48
+ debugger attachment changes; sensitive fields are refused; consequential actions require fresh
49
+ human approval and are revalidated; actions require an attended run; page content remains untrusted.
50
+ The Brave tool list does not include the App-specific `browser_upload`. WebKit remains an independent
51
+ backend with its existing narrower capabilities.
52
+
53
+ ## Consequences
54
+
55
+ - Harness and terminal settings can select isolated or Brave, pair, inspect status and stop the
56
+ connection. Hosts may disable the generic browser provider with `browser: false` and retain the
57
+ Harness bridge through `HarnessOptions.browserExtension`.
58
+ - The App keeps WebKit as its native backend and obtains shared Harness browser tools only when
59
+ Brave is selected. The App's connected-apps broker does not expose duplicate browser tools in
60
+ Brave mode.
61
+ - Pairing state lives in owner-only `<state dir>/browser/extension-bridge.json`; backend and
62
+ network settings remain in `<state dir>/browser.json`.
63
+ - The extension requires manual loading and site permissions. It does not provide a remote browser
64
+ service, history, cookie export, file uploads or general DevTools console/network access.
65
+ - D-25 remains open for selector-based operations in the App's WebKit Rust implementation and its
66
+ App-only upload contract. D-26 is partly resolved: Brave implements the shared ref tools, but the
67
+ App's Rust capability response is not yet backend-aware.
68
+ - The opt-in live acceptance in `src/browser/extension/brave-live.acceptance.test.ts` passed with a
69
+ disposable Brave profile and a loopback fixture. It covers pairing, exact-origin gating, snapshots,
70
+ actions, pause/stop and disconnect refusal; it does not exercise Chrome's optional-host grant
71
+ prompt for a non-loopback site. Use only disposable profiles for runtime verification; never use the
72
+ everyday profile.
package/docs/parity.md CHANGED
@@ -22,7 +22,8 @@ implemented is not a parity claim; acceptance is tracked per command.
22
22
  | Capability | Harness owner | Method or tool | Terminal surface | App surface | Status | Gap |
23
23
  |---|---|---|---|---|---|---|
24
24
  | Sessions and turns | `sessions/` | `sessions.run`, `terminal.*` | the terminal itself | Solo chat | parity | — |
25
- | Plan mode | `sessions/planMode.ts` | `plan: true` | `/plan`, Shift+Tab, plan card; shared outcome and bounded repair; plan/todo context restored after compaction | plan request, PlanCard; shared Harness classification and checkpoint behavior | ADR-0058 clarification contract approved; implementation and surface acceptance not verified | — |
25
+ | Pending follow-up and steering queues | Pi session queue, terminal queue projection | `followUp` and `steer` delivery modes | clickable native queue rows; `/follow-up` edits, sends next or removes, preserving images; on explicit user stop send one unconsumed steer before follow-ups; never drain on retry/failure/non-user interruption | inline Send now/Remove controls; successful completion stays FIFO, explicit stop prioritizes an undelivered steer; `/follow-up` points to the inline queue | implemented; acceptance not recertified | — |
26
+ | Plan mode | `sessions/planMode.ts` | `plan: true` | `/plan`, Shift+Tab, plan card; shared outcome and bounded repair; a reply that needs no plan ends with no card or warning; plan/todo context restored after compaction | plan request, PlanCard for a complete block and nothing otherwise; shared Harness classification and checkpoint behavior | ADR-0077 `no_plan` outcome implemented; surface acceptance not recertified | — |
26
27
  | Goal (`/goal`, ADR-0072) | `sessions/goal.ts`, `sessions/goalLoop.ts`, `sessions/goalPrompts.ts` | `sessions.run { goal }`, `goalMode` on the run record, model tools `get_goal`/`update_goal` while a goal exists, `extension: 'goal'` notices | `/goal [<objective>\|edit\|pause\|resume\|clear]`, `--rounds`/`--hours`, footer label, Esc pauses, goal restored `paused` from the Pi session | not wired: no goal control or status | terminal implemented; App presentation planned | PAR-017 |
27
28
  | Capability discovery | `tools/getCapabilities.ts` | read-only `get_capabilities` | terminal extension reports Pi's active tools and schemas; shared plan allowlist includes it | structured Harness run uses the same tool; no duplicate App-owned tool | implemented; acceptance not recertified | — |
28
29
  | Bash command mode | `sessions/sessions.ts`, `sessions/initialBash.ts` | `initialBashCommand`; Bash `workspaceWriteScope` | `!command` defaults to workspace; `!outside command` asks before running once | same `!command` and `!outside command` composer semantics | parity | — |
@@ -30,12 +31,12 @@ implemented is not a parity claim; acceptance is tracked per command.
30
31
  | Side questions | `sessions/sideQuestions.ts` | `sideQuestions.start/read/interrupt`; terminal `terminal.sideQuestions.*` | `/btw` contextual dock; the answer renders as Markdown, with scroll and copy | side-question dock through the shared service | parity | — |
31
32
  | Approvals and standing approvals | `sessions/approvals.ts` | `approvals.*`; run events `approval.requested`/`approval.resolved` | decision dock, `/permissions`; the call pauses in `waiting_for_approval` and resumes in the same turn | approvals dock above the composer, approval notification in the notification inbox and as a system notification while the window is hidden; settings | parity | — |
32
33
  | Opening a link | `core/links.ts` | `links.open`; `HarnessOptions.linkOpener` (ADR-0047) | hover shows the hand pointer and recolours the link, Ctrl+click (the primary click on macOS, whose ⌘ the mouse protocol cannot report) opens web, `mailto:`, file and folder targets with the system handler, a refused link posts a notice | ⌘-click on macOS and Ctrl+click elsewhere opens with the system handler, a plain click keeps the sidebar and falls back to the system for what it cannot show, `a:hover` uses `--link-hover` | implemented; acceptance not certified | — |
33
- | Questions (`ask_user`) | `tools/askUser.ts` | shared `ask_user` tool | plan-mode dock: strict aligned columns, bounded body scroll, 80×24 source sketch viewport; connected structural projection with high-confidence boundary normalization (PROP-0026, PROP-0028) | plan-mode form accepts the same 80×24 ASCII source contract with literal preview and accessible overflow; no typography change required | existing question surface parity; ADR-0058 plan clarification contract not verified | — |
34
+ | Questions (`ask_user`) | `tools/askUser.ts` | shared `ask_user` tool | plan-mode dock: strict aligned columns, bounded body scroll, fixed 80-column padded sketch canvas with a viewport-bounded frame and horizontal panning; connected structural projection with high-confidence boundary normalization (PROP-0026, PROP-0028, ADR-0080) | plan-mode form accepts the same permissive 80×24 ASCII source contract with a literal, horizontally scrollable preview; no App code change required | existing question surface parity; ADR-0058 plan clarification contract not verified | — |
34
35
  | Subagents | `sessions/subagentService.ts` | `spawn_subagent` (requires `title`, PROP-0030; `continue` resumes a stopped child, ADR-0043), `subagents.*`; plan mode allows only named eligible read-only research profiles; per-profile and total child budgets enforced in `Subagents.spawn` (ADR-0052), with the same refusal result in both surfaces | `/subagents` page with the `main` row pinned first, and live rows under the status line (mark, profile, parent-set title, `n of m tasks`), both rebuilt from saved runs when a conversation reopens; plan-mode research delegation uses the same Harness gate | reply trajectory; plan-mode research delegation uses the same Harness gate | parity | — |
35
36
  | Agent profiles | `agents/profiles.ts` | `spawn_subagent { agent }`, `agents.list/get/create/remove`; terminal `terminal.agents.*`; package profiles through `Loading.profilePaths` | `/agents` page: browse, create, edit in `$EDITOR`, copy, delete; package profiles listed under "From a plugin" and switched on `/plugins` | Solo `/agents` lists profiles; profile details and management are not wired | partial | PAR-010 |
36
37
  | Run todo list | `tools/todo.ts` | `todo_write`, `runs.todo` (`runs.checklist` deprecated); terminal `todo` telemetry | registered in terminal sessions and their subagents; each write shows as a todo block in the transcript; `/todo` shows the current list; a subagent row shows its `n of m tasks` | `/todo` shows the current list on demand; each write is not rendered inline | terminal done; App shows it on demand | PAR-001 |
37
38
  | Continuation of a stopped run | `sessions/continuation.ts` | `continueFrom`, `runs.continuation` | not offered (`/resume` reopens history only) | not offered | planned | PAR-003 |
38
- | Bounded replies and checkpoints (PROP-0041) | `sessions/workSlice.ts`, `tools/checkpoint.ts`, `sessions/runner.ts`, `terminal/stallWatch.ts` | `checkpoint_write`, `workSlice`, `runs.continuation.checkpoint`, `checkpoint`/`heartbeat` events | guard on every terminal reply with the shared tool; every live reply heartbeats and a reply with no progress for fifteen minutes ends as `stalled`; the handoff is recorded on the run; no checkpoint card or slice control yet | not wired | implemented; terminal card, App presentation and `/slices` open | PAR-004 |
39
+ | Bounded replies and checkpoints (PROP-0041) | `sessions/workSlice.ts`, `tools/checkpoint.ts`, `sessions/runner.ts`, `terminal/stallWatch.ts` | `checkpoint_write`, `workSlice`, `runs.continuation.checkpoint`, `checkpoint`/`heartbeat` events | guard on every terminal reply with the shared tool; every live reply heartbeats and a reply with no progress for fifteen minutes ends as `stalled`, a reply waiting on subagents is judged by each child's own clock, and the working row names what it waits on; the handoff is recorded on the run; no checkpoint card or slice control yet | not wired | implemented; terminal card, App presentation and `/slices` open | PAR-004 |
39
40
  | Accounts | `accounts/` | `accounts.*` | `/accounts` | Accounts | parity | — |
40
41
  | Models and effort | `accounts/choices.ts` | — | `/model`, `/effort` | composer pickers | parity | — |
41
42
  | Skills | `plugins/`, `tools/dispatch.ts`, `prompt/compile.ts` | `skill_load`; `Launch.requiredSkills` | `$name` at any word start (required for that request), `/skill:name`, browser Skills tab | composer `$` chip with `requiredSkills`; Capability | implemented; install routes taught in the release-owned skill, foreign manifest skills load as package skills; acceptance not recertified | — |
@@ -43,7 +44,7 @@ implemented is not a parity claim; acceptance is tracked per command.
43
44
  | Plugin authoring and proposal review | `plugins/proposals.ts`, `tools/defaultSkills.ts` | `plugin_propose`, `plugins.proposals.*` | `/plugins` Proposals tab: preview, accept or reject | structured model tool and skill; review UI not implemented | partial; App review pending | PAR-017 |
44
45
  | Connected apps (MCP) | `mcp/` | `connections.*`, broker tools | resource browser Connections tab | Connections | parity | — |
45
46
  | Web search and page reading | `web/` | `web_search`, `web_fetch` | world tools | world tools | parity | — |
46
- | Isolated browser | `browser/` | shared `browser_*` schemas | isolated world tools, headed on request, `/settings` Browser tab | native WebKit: navigate, tabs, visible-text read, page screenshot and page scroll without refs; unsupported operations refuse explicitly; live screenshot acceptance pending | partial; native backend gap | PAR-015 |
47
+ | Browser (isolated or Brave) | `browser/`, `browser/extension/` | shared `browser_*` schemas; `browser.config`, `browser.extension.*` | isolated or connected Brave tools; `/settings` selects the device backend, pairs the extension and shows status | WebKit: navigate, tabs, visible-text read, page screenshot and page scroll without refs; Brave: shared snapshot-ref tools through the paired extension; `browser_upload` remains WebKit-only; no fallback on disconnect | implemented; disposable loopback acceptance passed; non-loopback host grant prompt untested; WebKit gap remains | PAR-015 |
47
48
  | Web annotations | `annotations/` (PROP-0017, PROP-0018, PROP-0038, PROP-0048) | `annotations.*`, `annotations.capture.*`, `page_notes_*` | `/annotate:` page (Annotate and All annotations tabs; DevTools app list and launch), `/annotate:notes`, `/annotate:submit`, `/annotate:clear` | website tab Annotate panel (inspector, queue to the focused thread, app windows through `annotations.capture.*`) | App behind PROP-0048 (queue methods removed); native WKWebView gestures unaccepted | PAR-011 |
48
49
  | Managed worktrees | `worktrees/` (PROP-0016) | `worktrees.*` | `--worktree`, `/worktree`, `xagent worktree` | Solo `/worktree` lists records only; Solo starts sessions in the project folder and has no worktree level | partial | PAR-012 |
49
50
  | Wiki notes (model tools) | `wiki/`, `tools/` | `wiki_*` | `Terminals.ownTools` registers Wiki-read tools through the session socket; `xagent-wiki` remains a person-facing shell utility | structured Harness tools | implemented on both; acceptance not recertified | PAR-002: acceptance not recertified |
@@ -88,7 +89,8 @@ this fixed table.
88
89
  | `/goal` | Set or view a goal; edit, pause, resume, clear | Not wired: PAR-017 | Not certified per command |
89
90
  | `/btw` (`/side`) | Contextual side-question dock | Shared side-question service; bare form dismisses the dock | Not certified per command |
90
91
  | `/new` | New conversation | Replaces focused tab with a draft; archives its saved session | Not certified per command |
91
- | `/stop` | Interrupt current reply | Stops the focused session's reply | Not certified per command |
92
+ | `/stop` | Interrupt current reply; explicitly releases the next eligible native queue item after settlement | Stops the focused session and sends the next eligible queued item after an explicit stop | Not certified per command |
93
+ | `/follow-up` | Opens the follow-up and steering queue editor | Points to the existing inline queue and its Send now/Remove controls | Not certified per command |
92
94
  | `/session` (`/resume`) | Search/select native saved history in the current folder | Lists App Solo threads by folder and effective account; `all` widens only within that App list | PAR-008: separate catalogs; no native-history path opening; not certified |
93
95
  | `/history` (`/tree`) | Navigate branches in the current native session | Calls `session.navigate` with stable native entry ID; same-session navigation, optional summary, no fork fallback | Route uses Harness `sessions.navigate`; per-command acceptance pending |
94
96
  | `/fork` | Fork from a user message | Uses session tree/fork and opens the new session | Not certified per command |
@@ -18,6 +18,15 @@ Unstable preview of the current Harness work. Install with `npm install @xingula
18
18
 
19
19
  Test suites were not run for this candidate, at the request to bypass tests.
20
20
 
21
- Prepared from commit a9c25b3. Not published in the session that prepared it: npm
22
- authentication was unavailable (the registry answered 401). Follow `docs/releases/README.md` before
23
- publishing.
21
+ Prepared from commit f7c6fce. The candidate was built from that clean commit: SHA-256
22
+ `c58bbc67a22778f2724a9cbbc101a1bbbcd60434b05acab3013a3690a5c61fef`, npm integrity
23
+ `sha512-PJLxG42u+Wq8AliaplF93kbQeC2vZiqUzPW4pVXDIViVG6AwrGxzu6EvPEE5dZRAysSD07Aw3cLzvnEwLMgfrQ==`.
24
+ Published from commit f7c6fce with a manual `npm publish --access public --tag unstable`, after the
25
+ candidate was built from the same commit (SHA-256
26
+ `c58bbc67a22778f2724a9cbbc101a1bbbcd60434b05acab3013a3690a5c61fef`, tarball shasum
27
+ `e372cb575bffe266ac40d0d5c58581b7c45db1f4`, npm integrity
28
+ `sha512-PJLxG42u+Wq8AliaplF93kbQeC2vZiqUzPW4pVXDIViVG6AwrGxzu6EvPEE5dZRAysSD07Aw3cLzvnEwLMgfrQ==`).
29
+ The registry held the version in its malware scan for about 27 minutes before `unstable` moved to
30
+ it. Because the suites and the exact-tarball smoke checks did not run, and the workflow's OIDC path
31
+ was not used, this version carries no provenance attestation and `scripts/release-verify.mjs`
32
+ cannot verify it. The release workflow's gates are unchanged for the next release.
@@ -0,0 +1,24 @@
1
+ # Xingular Harness 1.5.1-unstable
2
+
3
+ Unstable preview of the current Harness work. Install with `npm install @xingular/harness@unstable`.
4
+
5
+ - **Tool names in the terminal transcript.** Tools render in CamelCase (`todo_write` becomes
6
+ `TodoWrite`, `computer_screenshot` becomes `ComputerScreenshot`) instead of the raw tool id. The
7
+ todo card reads `Todo`, a custom message carries its type in the same form, a running child's
8
+ activity reads `Read` rather than `read`, and the phone page's transcript and approval card use
9
+ the same names.
10
+ - **Claude Code prompt study.** `STUDY-0007` reads Claude Code v2.1.88's shipped prompt from the
11
+ leaked source map and compares it with this harness's pack in 20 areas;
12
+ `scripts/study-claude-prompt.mjs` reproduces the measurement.
13
+
14
+ Test suites were not run for this candidate, at the request to bypass tests.
15
+
16
+ The candidate was built from commit `f6870fb`: SHA-256
17
+ `a9f49f244c160cf9bf9e6cf4192822ca658cba24f055c9196ea8a546bf5e45bd`, npm integrity
18
+ `sha512-uWzJE+QJfmsgOuS+EHAc2hbAt6lvHmvzhVqkyAQfwj9rwVb+kkZXnzBk94qHATI17pS3mX8IyJjNePTYJFTsXg==`.
19
+ Publication is blocked on npm's one-time password: publish the tarball with the browser flow, or push
20
+ the `v1.5.1-unstable` tag and let the release workflow publish it with provenance.
21
+
22
+ The tree has since moved past `f6870fb` (the transcript names reach the todo card, custom message
23
+ types, a child's activity and the phone page), so that tarball predates part of the note above: build
24
+ a new candidate from the current commit before publishing.
@@ -0,0 +1,59 @@
1
+ # Xingular Harness 1.6.0-unstable
2
+
3
+ Unstable preview of the current Harness work. Install with `npm install @xingular/harness@unstable`.
4
+
5
+ - **Your own signed-in Brave profile (ADR-0082).** `/settings` → Browser now chooses the browser
6
+ backend: the isolated Harness profile, or Brave through a paired unpacked extension that ships in
7
+ the package. Pairing uses a one-time `port:code` over an authenticated loopback bridge; each site
8
+ is allowed separately by exact origin in the extension popup; an in-page badge offers Pause and
9
+ Stop. The agent gets the same `browser_*` tools and the same approval gate, snapshot refs
10
+ included, and `browser_upload` stays WebKit-only. A disconnect refuses calls and never falls back
11
+ to the isolated browser. New methods `browser.config`, `browser.configure` and
12
+ `browser.extension.status|start|stop`; `browser.json` gains `backend`.
13
+ - **A reply that waits on children or on a flaky network no longer looks stalled (ADR-0078).** Each
14
+ subagent the reply waits on gets its own stall clock, the parent's clock pauses while it only
15
+ waits, and a transient model failure retries for about four minutes instead of fourteen seconds.
16
+ The working row says what the reply is doing: waiting on children, running a long tool, waiting
17
+ for the model, reconnecting, or making no progress.
18
+ - **A plan run may conclude that no plan is needed (ADR-0077).** A plan reply with no plan block
19
+ ends as `no_plan` with no card, notice or repair turn. Only a block the model started and left
20
+ incomplete is repaired, once, in the same run.
21
+ - **No work-slice budget by default (ADR-0081).** A reply is no longer ended at 320 tool calls or
22
+ 60 minutes and continued as a "next slice". `workSlice` on a run and `XINGULAR_WORK_SLICE` still
23
+ set a budget, and the three-compaction guard stays.
24
+ - **Phase-led, flexible reporting (ADR-0079).** The `reporting` guidance is written for a reader
25
+ picking up cold: verification that did not run says so, subagents get a concise caller-facing
26
+ report contract, and a substantial report may use content-named headings. The duplicated
27
+ progress-update line moved out of `reporting` and stays in `writing`.
28
+ - **The follow-up and steering queue has a page.** `/follow-up`, and clicking a `Steer:` or
29
+ `Follow-up:` row, opens an editor for the queued messages with edit, send next, remove and
30
+ return. An explicit stop now sends one unconsumed steering message before the follow-ups.
31
+ - **Question sketches use one 80-column canvas (ADR-0080).** The model draws on a fixed 80-column
32
+ canvas and the terminal pads every valid sketch to it, with the frame inside the dock and
33
+ horizontally pannable. Stored sketches and answer payloads are unchanged.
34
+ - **CamelCase names everywhere in the transcript.** The 1.5.1 candidate predated the last of this
35
+ work, so this release carries the todo card, custom message types, a child's activity and the
36
+ phone page using the same `TodoWrite`/`ComputerScreenshot` form.
37
+
38
+ Two prompt sections change (`reporting` and the work-slice wording), so the stable prefix misses the
39
+ provider cache once per existing session after the upgrade. The Brave backend is opt-in and off by
40
+ default; the isolated browser is unchanged.
41
+
42
+ The release workflow builds the tarball from the `v1.6.0-unstable` tag, reruns CI on Ubuntu and
43
+ macOS, smoke-tests that exact tarball on both, and attaches it with `SHA256SUMS` to the GitHub
44
+ release before publishing it to npm's `unstable` dist-tag with provenance. The published SHA-256 and
45
+ npm integrity are the ones in that attached file, which `scripts/release-verify.mjs` checks against
46
+ the registry.
47
+
48
+ Checks on this source before tagging: `npm run typecheck` clean; `npm test` green (220 test files,
49
+ 1,794 passing tests, 4 skipped); `npm run prove` green; `npm run release:smoke -- <tarball>
50
+ --terminal` green in a fresh installation (SDK, subpaths, Doctor, packaged CLI, JSON-lines server
51
+ and the terminal in an isolated PTY); `npm run release:preflight` reporting `1.6.0-unstable` unused,
52
+ with `unstable` on `1.5.1-unstable` and `latest` on `1.2.0`. The tarball carries 880 files and the
53
+ `browser-extension/` folder.
54
+
55
+ The live Brave acceptance test did not run: it needs `XINGULAR_BRAVE_LIVE=1` and a disposable Brave
56
+ window, so live control of a real profile is unverified here. The App's own checks were run against
57
+ this Harness source and are reported separately: its typecheck reports pre-existing `@xingular/ui`
58
+ prop-type errors (the `Xingular UI` path its tsconfig expects is absent), and one pre-existing Solo
59
+ test fails because the App does not wire `/goal` (PAR-017). Neither comes from this change.
@@ -29,6 +29,22 @@ written and the claim moves to checked behavior. Cite material as
29
29
  - [How peers shape the final report](STUDY-0006-reporting-prompts.md) — the final-answer rules of
30
30
  Claude Code, Codex and Kimi; what this harness's 3,400-character reporting section lacked; and the
31
31
  large content-block prompt in [ADR-0073](../decisions/ADR-0073-large-reporting-prompt.md).
32
+ - [Claude Code's prompt against this harness's, in 20 areas](STUDY-0007-claude-code-prompt.md) — the
33
+ leaked v2.1.88 prompt measured against the pack a real session was given: 12.9k characters against
34
+ 23.4k, the 9.2k reporting section with no Claude Code counterpart, and the rules Claude Code has
35
+ that the harness does not (parallel calls, read-before-edit, minimal diffs).
36
+ - [How Codex and Claude Code use the person's browser](STUDY-0008-browser-control.md) — Claude Code drives
37
+ the person's own Chrome as a built-in MCP server with refs, console, network and JavaScript; Codex has
38
+ no browser in the CLI and reaches the person's browser only through Computer Use and its extension,
39
+ keeping a separate-profile built-in browser of its own; this harness now offers isolated Chromium,
40
+ App WebKit, and an explicitly paired Brave extension through shared browser tools. The Brave bridge
41
+ is implemented with live disposable-profile acceptance still pending; see [ADR-0082](../decisions/ADR-0082-connected-brave-browser.md).
42
+ - [Claude Code's report structure against this harness's, phase by phase](STUDY-0009-claude-code-reporting.md)
43
+ — the leaked v2.1.88 external and internal prompts plus the published Concise style against the
44
+ original 3,696-character reporting section: audience, pre-tool text, in-flight updates, first
45
+ sentence, change content, evidence, next steps, headers, tables, length, references and examples.
46
+ Neither Claude Code build requires headers; ADR-0079 records the Harness's choice to encourage
47
+ content-named `#` / `##` headings in substantial reports.
32
48
  - [Does the harness tell an agent how to install a skill](STUDY-0005-skill-installation-guidance.md)
33
49
  — the global and project routes this harness already loads, checked with its own resolver; why a
34
50
  session model is told none of them; and how Codex, Gemini, Claude Code, Kimi and others put that
@@ -48,7 +64,7 @@ Clones live in `external/` (this repo). Field notes in `sources/`.
48
64
  | openhands | [All-Hands-AI/OpenHands](https://github.com/All-Hands-AI/OpenHands) (UI) + [OpenHands/software-agent-sdk](https://github.com/OpenHands/software-agent-sdk) (agent server runtime) | [openhands.md](sources/openhands.md) | inventory |
49
65
  | kimi-cli | [MoonshotAI/Kimi-CLI](https://github.com/MoonshotAI/Kimi-CLI) | [kimi-cli.md](sources/kimi-cli.md) | inventory |
50
66
  | smolagents | [huggingface/smolagents](https://github.com/huggingface/smolagents) — CodeAct | [smolagents.md](sources/smolagents.md) | inventory |
51
- | claude-code | [Anthropic Claude Code](https://www.anthropic.com/claude-code) — not open source; audit npm bundle, published prompts, observed behavior | [claude-code.md](sources/claude-code.md) | no clone |
67
+ | claude-code | [Anthropic Claude Code](https://www.anthropic.com/claude-code) — not open source; the leaked v2.1.88 source map gives a full read | [claude-code.md](sources/claude-code.md) | v2.1.88 source in `external/claude-code-src`, prompt audited |
52
68
 
53
69
  ## Arms
54
70