@aragon-agent/cli 0.6.34 → 0.6.36

package/CHANGELOG.md

@@ -8,6 +8,27 @@ were written.
 ## Unreleased
 
 ### Added
+- **Acceptance criteria: a plan now has to say how you would know it is done.**
+  `todo_write` takes a second array in the SAME call - one checkable statement per
+  line plus the check the model will run for it - so the steps and the criteria can
+  never drift apart. When every step is completed the plan enters a derived
+  **verify** phase: the model checks one criterion at a time and must record what it
+  actually saw, and a verdict without evidence drops back to unchecked. A verdict is
+  never undone by quietly omitting the criterion from the next list; re-checking one
+  takes `reopen: true` or `/todo acceptance reset`. The rail grows a second section
+  (`ACCEPT 2/5` while working, `VERIFY 2/5` once the steps are done, one `AC 2/5`
+  line on a short terminal), the transcript card lists the failed criteria with their
+  evidence, `/todo status` repeats the table, the status line carries `Accept 2/5`
+  when the rail cannot, `-p` prints one `[accept] 1/3 checked` line per transition,
+  and `aragon exec`'s `todo`/`result` events carry the counts (never the evidence).
+  "Every step is completed" is no longer the same claim as "the work is done": a new
+  `planIsResolved` judge decides the follow-through, the grace-window re-check and
+  the new-task clear together, and auto-continue counts a settled criterion as
+  progress so a plan is not abandoned after its second check. New `todo.acceptance`
+  key (`--todo-acceptance` / `--no-todo-acceptance` / `ARAGON_TODO_ACCEPTANCE` /
+  `/todo acceptance on|off|reset`, `todo-acceptance` capability); with it off every
+  surface is byte-identical to before. See
+  [Acceptance criteria](README.md#acceptance-criteria).
 - **Background jobs, and a bounded wait for them.** A command that finishes on its
   own but takes far longer than one tool call - `python train.py`,
   `huggingface-cli download`, `cargo build --release` - can now be started with

package/README.md

@@ -23,7 +23,7 @@ full-screen mode (`--scroll-mode app`), where the application owns the mouse.
 **On Windows with Node < 22.17.0 (or 24.0/24.1), aragon turns the missing
 console flag on itself at startup**, which restores `Shift+Tab` on most
 machines. Where that is blocked — a locked-down PowerShell, an old console
-host — **`Ctrl+P` toggles the mode instead**; it is bound on every platform and
+host — **`Ctrl+N` toggles the mode instead**; it is bound on every platform and
 does not depend on the console mode, the Node version, or anything else.
 Everything else — the agent loop, every tool, skills, teams, `aragon -p` — works
 exactly the same on Node 18 and 20, which is why `engines` stays at `>=18`
@@ -44,7 +44,7 @@ on. When that works, `Shift+Tab` arrives as `CSI Z` again and mouse reporting
 goes back on with it — the wheel should follow by the same route, though only
 the key has been measured. When it does not work, aragon says so at startup and
 turns mouse reporting off rather than asking for reports that can never arrive;
-`Ctrl+P` and `/plan` switch mode, and `Shift+↑` / `Shift+↓` scroll.
+`Ctrl+N` and `/plan` switch mode, and `Shift+↑` / `Shift+↓` scroll.
 
 ## Install
 
@@ -389,6 +389,7 @@ below works there too, in either position — `aragon --model m exec "…"` and
 | `--todo` / `--no-todo` | Enable or disable todo planning. `--no-todo` does not register the `todo_write` tool at all |
 | `--todo-panel` / `--no-todo-panel` | Show or hide the right-hand todo rail. The tool stays registered either way |
 | `--todo-follow <mode>` | What happens when a run ends with steps left: `notify` (default), `auto`, `off` |
+| `--todo-acceptance` / `--no-todo-acceptance` | Plan acceptance criteria with every list (the default) or steps only. Persisted, so the pair exists to keep "not passed" distinct from `true` |
 | `--retry` / `--no-retry` | Retry a failed provider call with backoff (the default), or fail on the first error |
 | `--retry-max <n>` | Retries after the first attempt for this run, clamped to `[0, 20]`. `0` turns retry off |
 | `--fast` / `--no-fast` | Enable or disable the fast model tier. Off by default; see **Fast model tier** |
@@ -480,6 +481,68 @@ displayed. `aragon --resume` of such a file inherits the full context and shows
 empty screen until the next turn - the alternative, inventing entries, would put
 a conversation the user never saw onto their screen.
 
+## Agent kits
+
+A **kit** is a named preset that pins the three model roles at once: which
+profile runs the main model, which runs the fast tier (or none), and which model
+subagents use. It is a set of *references* to the profiles you already have, so
+editing a profile changes every kit that names it — which is what "preset"
+should mean.
+
+Kits exist in three layers, and only one of them writes a file:
+
+| Layer | Where | Lives until |
+| --- | --- | --- |
+| **Saved** | `/settings` → `Kit pack` → `New kit` / `e` | you delete it |
+| **Default** | `Kit pack` → `s` (*Set default*), then `Ctrl+S` | you change it |
+| **Session** | `Ctrl+K`, or `/kit use <name>` | this process exits |
+
+The first row of the `Ctrl+K` overlay is always `Current setup` — the bindings in
+`config.json` — and choosing it releases a session kit, so the feature has an exit
+as well as an entrance.
+
+```
+Agent kits                     Session only; set a default in /settings
+  [In use] Deep work                                    [Default]
+  Current setup
+  ────────────────────────────────────────────────────────
+  main      anthropic:claude-opus-4-6        default URL
+  subagent  anthropic:claude-haiku-4-5       profile Haiku
+  fast      anthropic:claude-haiku-4-5       tier on
+  ────────────────────────────────────────────────────────
+  Enter use for this session | Esc close
+```
+
+- **`Ctrl+K` never writes the config file.** That is a promise with a test
+  behind it: the session path runs the same prepare chain as a save and skips
+  only the commit, so a failed switch leaves both the file and the session
+  exactly as they were, and a successful one produces no bytes on disk.
+- **A save decides slots, not the whole kit.** Change the subagent model while a
+  session kit is in effect and only the subagent slot leaves the kit; main and
+  fast keep the kit's choices. Rebinding main releases main, and so on.
+- **The status bar names what is running** with a `kit <name>` chip, and appends
+  `~` when the live connections no longer match the preset (for example after
+  `/model`). With no kit in effect the chip is not rendered at all.
+- **`Ctrl+G` reveals `^K switch kit`** in the header row; on terminals narrower
+  than 79 columns it gives way to `^L` and `^I`, and `/help` plus `/kit` are the
+  fallback. (`Alt+K` is where `Ctrl+K`'s old "kill to end of line" went.)
+- **Setting a default applies its three slots**, because "the default" *is* the
+  binding in the config file. A pointer alone would not be in effect at the next
+  launch.
+- **Deleting a profile deletes the kits that name it**, in the same
+  confirmation, and says so: a preset pointing at a deleted profile can neither
+  be applied nor repaired from any screen. Kits that a hand-edited file leaves
+  dangling degrade to `Profile missing` and are refused with a stable message.
+
+Config keys (all inside the existing `modelProfiles` section, so an older build
+ignores them rather than rejecting the library):
+
+| Key | Meaning |
+| --- | --- |
+| `modelProfiles.kits` | The library: id, name, main profile, fast profile (`null` = off), subagent mode + profile |
+| `modelProfiles.defaultKitId` | The default kit, or `null`. A dangling value heals to `null` |
+| `modelProfiles.kitsVersion` | Migration marker, always `1`. **Downgrading drops kits** (the older build's whitelist does not know the keys); profiles and bindings are unaffected, and `New kit` prefills from your current setup so rebuilding is one Save |
+
 ## Slash commands (interactive)
 
 | Command | Action |
@@ -489,6 +552,7 @@ a conversation the user never saw onto their screen.
 | `/settings` | Open the settings screen |
 | `/thinking <level>` | Set the thinking level |
 | `/max-tokens [<n>\|auto\|default]` | Output token cap; with no argument, reports the setting **and** the effective cap for the live model |
+| `/agents` | Open the **subagent board**: every child of the current (or most recent) dispatch, its full facts, its result and its 24-event timeline. The same board as `Ctrl+B`. Read-only - it can never nudge, kill or re-dispatch a child |
 | `/tools` | List active tools |
 | `/clear` | Start a new conversation: interrupt the run, archive the old session, clear the context, transcript, queues and todo panel. `/clear screen` only clears the screen |
 | `/reset` | Alias of `/clear` - one implementation, so the two cannot drift apart |
@@ -498,13 +562,15 @@ a conversation the user never saw onto their screen.
 | `/session [copy\|pin]` | Report this session's id, file and branch parent; `copy` puts `aragon --resume <id>` on the clipboard; `pin` makes `sessions prune` skip the file forever |
 | `/plan [on\|off\|status]` | Toggle plan mode - the keyboard-free equivalent of `Shift+Tab` |
 | `/team [on\|off\|max <n>\|overseer on\|off\|model [main\|fast\|<profile-name>]]` | Team subagents: status and the live roster, the on/off switch, the fan-out width, the dispatch supervisor |
-| `/todo [on\|off\|panel on\|off\|follow <mode>\|clear\|continue]` | Todo planning: status, the on/off switch, the rail, follow-through, and picking an unfinished plan back up |
+| `/todo [on\|off\|panel on\|off\|follow <mode>\|acceptance on\|off\|reset\|clear\|continue]` | Todo planning: status, the on/off switch, the rail, follow-through, acceptance criteria, and picking an unfinished plan back up |
 | `/retry [show\|on\|off\|max <n>]` | API retry: the effective ladder, the on/off switch, the retry count. Takes effect **in the running session**, children included |
 | `/fast [on\|off\|model <id>\|provider <id>\|same\|review <n\|off>]` | Fast model tier: status and this session's totals, the on/off switch, which model, the review cadence |
 | `/compact [status\|on\|off\|threshold <n>\|keep <n>\|history\|show <n>\|<instructions>]` | Context compaction: occupancy and this session's totals, the on/off switch, when it fires, how much it keeps. With free text, compacts now and tells the summarizer what to pay attention to |
 | `/context` | What the context gauge is showing: occupancy split into measured and estimated, the window and **where it came from**, how stale the measurement is, compaction's actual state, and what the session-spend readout includes |
 | `/bg [list\|logs <id> [n]\|stop <id\|all>\|status]` | Background jobs (services and tasks): what is running, its log tail, and how to stop it. `list` is the default |
 | `/update [status\|now\|skip\|off]` | Auto-update: what is running, what is available, where it came from, when it looks again. `now` forces a check past the throttle; `skip` declines the offered version until a newer one appears |
+| `/prompts [manage\|list\|show <n\|title>\|add [title :: body]\|edit <n\|title>\|rm <n\|title> [--yes]\|move <n\|title> <to>\|favorite <n\|title>\|sort name\|recent\|favorite]` | Quick prompt library: `/prompts` opens the picker (Enter fills the draft, `Ctrl+Enter` fills and sends), `add`/`edit` open the editor. `<n\|title>` accepts a 1-based index, an exact title, or a unique title prefix; a repeated title is refused **with its candidates** rather than guessed |
+| `/kit [list\|use <name\|current>]` | Agent kits: what is in effect and where it came from, the whole library with each preset's three connections, or apply one for **this session**. `/kit use current` releases it. Never writes the config file |
 | `/theme <auto\|warm\|cool\|light>` | Switch the color theme live (persisted) |
 | `/expand [n]` | Expand / collapse the n-th-from-last (default last) tool card |
 | `/copy` | Copy the last answer to the clipboard |
@@ -526,8 +592,11 @@ completes, `Up` / `Down` moves the selection, `Esc` closes it.
 | `Enter` | 空闲时发送，运行时加入队列；收到 Core 已纳入会话的接收凭据后才移除 Queue 提示 |
 | `Ctrl+J` / `Alt+Enter` / `Shift+Enter` | 插入换行。应用默认请求终端区分修饰键（Windows 用 win32-input-mode，macOS/Linux 用 kitty/modifyOtherKeys），Shift+Enter 在支持的终端开箱即用；不支持的终端用 Ctrl+J。参见 [终端设置](#terminal-setup-for-shiftenter) |
 | `Shift+Tab` | Toggle **plan mode** (`BUILD` <-> `PLAN`). Equivalent: `/plan`. On Windows this needs a console flag Node < 22.17 does not set; aragon sets it at startup — see [Requirements](#requirements) |
-| `Ctrl+P` | The same toggle, over a channel no Windows console can swallow. Use it if `Shift+Tab` does nothing |
+| `Ctrl+N` | The same toggle, over a channel no Windows console can swallow. Use it if `Shift+Tab` does nothing. Written `Ctrl+N` everywhere, never `^N` - `^N` is the header's line-offset chip |
+| `Ctrl+P` | **Quick prompt library**: the manager - create, edit, delete, favourite and reorder your own prompts. See [Quick prompt library](#quick-prompt-library) |
+| `Ctrl+B` | **Agents board**: the details of a running dispatch. Left column lists every child (not just the five the roster fits), the right-hand sections show the selected child's state, turns, tools, tokens, files, mail, supervisor actions, retries, compaction, its current activity, its final answer or error, and the last 24 events. `Up` / `Down` change the child, `PgUp` / `PgDn` and the wheel page, `Esc` closes. The board survives the end of the dispatch - it freezes the last live snapshot and marks it `dispatch ended` - and opens after the fact from the session's own team card, where everything except the live-only sections (current activity, mail count, timeline) is still readable. Pressing `Ctrl+B` again, or `/agents`, opens and closes it |
 | `Ctrl+I` | **Build the project index** (bundled `project-indexer` skill). Asks for confirmation, then runs as a normal agent turn that writes `.claude-index/` and updates `CLAUDE.md`. On terminals without keyboard enhancement `Ctrl+I` is the Tab byte and completes input instead |
+| `Ctrl+K` | **Agent kits**: pick a model preset — main / subagent / fast profile in one choice — for **this session only**. Nothing is written to the config file. `Current setup` (the first row) releases the preset and returns to the file's bindings. See [Agent kits](#agent-kits) |
 | `Esc` | Close an overlay or popup; otherwise ask for interruption confirmation. While idle, cancel pending auto-continuation |
 | `Esc` ×2 | Interrupt the run when pressed within 1.5 seconds. Other keys or closing a menu cancel the first press |
 | `Esc` again after interruption | **Force-stop** a run that is still stopping, kill foreground shell commands and return to `idle`; background services remain running |
@@ -545,7 +614,7 @@ completes, `Up` / `Down` moves the selection, `Esc` closes it.
 | `Home` / `End`, `Ctrl+A` / `Ctrl+E` | Cursor to line start / end |
 | `Alt/Ctrl+←` / `→` | Word-wise cursor jump |
 | `Ctrl+W`, `Alt+Backspace` | Delete the previous word |
-| `Ctrl+U` / `Ctrl+K` | Kill to line start / end |
+| `Ctrl+U` / `Alt+K` | Kill to line start / end (`Alt+K` was `Ctrl+K`, which now opens the kit picker) |
 | `Up` / `Down` | 按视觉行移动；空草稿或正在浏览历史时，才在边界召回历史，避免覆盖新草稿 |
 | Paste | Inserted up to 6 lines / 400 chars; larger collapses to `[Pasted text #1 +15 lines]`, which deletes as one unit and expands in full when you send. Escape sequences are removed as sequences (so a coloured log arrives as its text, not as `[38;2;18;52;86m`), and a TAB is kept in the payload but laid out at the terminal's 8-column tab stops, so the draft and the echoed message line up. `--no-paste` restores the pre-0.6.3 behaviour |
 | `?` | Open help (empty input) |
@@ -586,6 +655,47 @@ Turn the whole thing off with `--no-mouse`, `ARAGON_MOUSE=0`, or
 nothing in full-screen mode — it does not fall back to editing your draft, which
 is what it used to do. Inline mode is unaffected either way.
 
+## Quick prompt library
+
+`↑` history remembers what you said. It does not remember what you say *every*
+day. The prompt library is a small ordered collection of prompts you write
+yourself - a title, a body, and optional tags, a note and a favourite flag -
+that you can drop into the composer instead of retyping.
+
+- **`Ctrl+P`** opens the manager. `Enter` edits the focused prompt, `a` creates
+  one, `d` deletes it (the confirmation page shows the body's size first), `f`
+  toggles favourite, and `Ctrl+`↑ / `Ctrl+`↓ move one prompt a position. The
+  order you see **is** the order on disk. In the editor, `Ctrl+S` saves and
+  `Tab` to the `Save` row then `Enter` does exactly the same thing - the second
+  path exists because a terminal can eat a `Ctrl` byte.
+- **`/prompts`** opens the picker. Type to filter (`clr` finds
+  `clean code review`); `↑` / `↓` choose; `Enter` **fills the draft** without
+  sending it; `Ctrl+Enter` fills and sends in one step; `1`-`9` pick directly
+  while the query is empty; `Ctrl+S` cycles `manual` / `recent` / `favorite`;
+  `Tab` switches to the manager.
+- **A fill never overwrites what you typed.** Into a non-empty draft it inserts
+  at the caret with a blank line between (`prompts.insertMode: 'replace'` does
+  overwrite if you prefer that), and it never becomes a `[Pasted text #1]`
+  token - an eight-line prompt arrives as eight lines.
+- **Where it lives.** `~/.aragon-agent/quick-prompts.json`, written `0600` on
+  POSIX - not `config.json`, which is where an API key lives. An unreadable file
+  is copied to `quick-prompts.json.bak` and reported once; individual records
+  that cannot be parsed are skipped **and counted** rather than silently
+  dropped in bulk.
+- **Where a key cannot be delivered.** `Ctrl+Enter` needs keyboard enhancement;
+  without it your terminal sends a plain `Enter`, so the prompt is filled and
+  you press `Enter` yourself. `Ctrl+`↑/↓ can lose its modifier on some
+  terminals - reordering is then always available as
+  `/prompts move <n> <to>`.
+
+| Config key | Default | Meaning |
+| --- | --- | --- |
+| `prompts.enabled` | `true` | `false` stops `Ctrl+P` and the header hint. `/prompts` stays registered and reports that the library is off |
+| `prompts.insertMode` | `insert` | What a fill does to a draft that already has text: `insert` keeps your words, `replace` overwrites them |
+| `prompts.hint` | `true` | Whether the header's right-hand column teaches `^P prompts` |
+
+`ARAGON_PROMPTS=0` is the same switch as `prompts.enabled: false`.
+
 ## Terminal setup for Shift+Enter
 
 应用现在会主动请求终端区分修饰键：Windows 上推送 win32-input-mode（conhost 与
@@ -1066,7 +1176,7 @@ it is an informed act rather than a stray keypress.
 
 ### If `Shift+Tab` does nothing
 
-**Press `Ctrl+P`.** It is the same toggle and it does not depend on any of the
+**Press `Ctrl+N`.** It is the same toggle and it does not depend on any of the
 machinery below.
 
 To find out why, run this in the same terminal, press `Shift+Tab` once, then
@@ -1196,6 +1306,58 @@ planning discipline and hides the column — which is what you want with a scree
 reader, and what `--no-todo-panel` does for a whole session. Under `-p` the plan
 is written to stderr as `[todo] 3/7 <step>` lines, suppressed by `--quiet`.
 
+### Acceptance criteria
+
+A plan is a promise; the criteria are the evidence. Every plan therefore carries a
+second array — **how you will know each part is really done**, one checkable
+statement per line, each with the check the model intends to run:
+
+```jsonc
+{ "todos": [ /* … */ ], "acceptance": [
+  { "content": "the rail shows 3 items on an 80x24 terminal", "check": "resize and look" },
+  { "content": "npm test -- runs green", "check": "npm test" } ] }
+```
+
+They arrive in the **same** `todo_write` call — one tool, one writer, one full
+replacement — so they cannot drift from the steps. Once every step is completed the
+plan enters **VERIFY**: the model checks one criterion at a time, and a verdict must
+name what it actually saw (`evidence`). A verdict without evidence drops back to
+unchecked. A verdict is *not* undone by quietly leaving the criterion out of the next
+list; re-checking one takes `reopen: true` (or `/todo acceptance reset`).
+
+The rail grows a second section under the steps — `ACCEPT 2/5` while working,
+`VERIFY 2/5` once every step is done, folding to a single `AC 2/5` line when the
+terminal is short:
+
+```
+  transcript ...                            | TODO             7/7
+                                            | ####------
+                                            |
+                                            |    all steps complete
+                                            | VERIFY           2/5 !1
+                                            | x npm test -- runs green
+                                            | ! no console errors on startup
+                                            | > README documents it
+```
+
+The failures are the point: the transcript card lists each failed criterion with the
+evidence under it, `/todo status` repeats the table on demand, and the status line
+carries `Accept 2/5` when the rail cannot. `-p` prints one `[accept] 1/3 checked`
+line per transition, and `aragon exec`'s `todo` and `result` events carry the same
+counts (never the evidence).
+
+```
+/todo acceptance            show the criteria and every verdict so far
+/todo acceptance on|off     plan criteria at all (persisted; default on)
+/todo acceptance reset      send every verdict back to unchecked
+--todo-acceptance           / --no-todo-acceptance, ARAGON_TODO_ACCEPTANCE
+todo.acceptance             true | false in config.json
+```
+
+With `todo.acceptance: false` every surface — the prompt block, the tool's arguments
+and result text, the rail, the card, the `[accept]` lines and the exec fields — is
+byte-identical to a build without this feature, and a plan needs no criteria at all.
+
 右栏从 76 列开始显示，正文至少保留 62 列；100 列终端使用 15 列右栏。
 已有任务时，团队面板按实际行数占用空间，补全菜单的边框和提示也计入预算。
 例如 100×20、团队占 8 行时，TODO 仍以 4 行紧凑视图显示。
@@ -2433,6 +2595,7 @@ env / `.env` → CLI flags**.
   `ARAGON_TEAM` (`0` disables team subagents), `ARAGON_TEAM_MAX` (fan-out width),
   `ARAGON_TODO` (`0` disables todo planning entirely),
   `ARAGON_TODO_FOLLOW` (`notify` / `auto` / `off`),
+  `ARAGON_TODO_ACCEPTANCE` (`0` plans steps only: no criteria, no verify phase),
   `ARAGON_RETRY` (`0` disables API retry), `ARAGON_RETRY_MAX` (retry count; `0`
   also disables),
   `ARAGON_UPDATE` (`0` / `off` disables auto-update, `notify` reports without
@@ -2441,7 +2604,9 @@ env / `.env` → CLI flags**.
   `ARAGON_SKILLS_DISABLED` (comma-separated names), `ARAGON_LOG_LEVEL`,
   `ARAGON_LOG_FILE` (`0` disables), `ARAGON_LOG_DIR`,
   `ARAGON_TRANSCRIPT_RETAIN`, `ARAGON_RENDER_GOVERNOR` (`0` disables),
-  `ARAGON_MAX_RENDER_INTERVAL_MS`, `ARAGON_SCROLL_MODE` (`native` / `app`).
+  `ARAGON_MAX_RENDER_INTERVAL_MS`, `ARAGON_SCROLL_MODE` (`native` / `app`),
+  `ARAGON_PROMPTS` (`0` disables the quick prompt library: no `Ctrl+P`, no
+  header hint; `/prompts` still reports that it is off).
   `ARAGON_HOME` is the exception that **cannot** come from `.env`: it is read
   before any `.env` is loaded. See [Files & logs](#files--logs).
 - **User config file** (`aragon config path`): JSON, written `0600` on POSIX.
@@ -2482,6 +2647,7 @@ env / `.env` → CLI flags**.
 | `todo.enabled` | `true` | Master switch for todo planning. `false` means the `todo_write` tool is never registered — `/todo on` in such a session saves the setting for next launch and says so, rather than advertising a tool that is not there. |
 | `todo.panel` | `true` | Render the right-hand rail. Independent of `todo.enabled` on purpose: a screen-reader user wants the planning discipline without the column, and the system prompt varies one sentence accordingly. |
 | `todo.followThrough` | `"notify"` | What happens when a run ends with steps left. `notify` says so and stops; `auto` continues the plan after a 3-second grace window, bounded by two structural limits (one fruitless attempt, 25 continuations per plan) that are deliberately **not** settings; `off` is silent. An unrecognized value falls back to `notify`. |
+| `todo.acceptance` | `true` | Plan acceptance criteria with every list, and show them (`ACCEPT` / `VERIFY` in the rail, the card's second block, `Accept 2/5` on the status line). `false` plans steps only: no criteria, no verify phase, and every surface is byte-identical to a build without the feature. `/todo acceptance off` changes it for the running session too. |
 | `retry.enabled` | `true` | Retry a failed provider call at all. Unlike `team.enabled` / `todo.enabled` this decides nothing at construction — the policy lives on the provider registry, so `/retry on` / `/retry off` takes effect **in the running session**, subagents included. |
 | `retry.maxRetries` | `10` | Retries **after** the first attempt, clamped to `[0, 20]`. `0` is the kill switch and is a real value here, not a typo for the default. The 20 is a hard ceiling enforced twice — on the config value and again inside the engine — so hand-editing this file cannot raise it. |
 | `retry.initialDelayMs` | `1000` | The first wait, clamped to `[100, 30000]`. |

package/dist/agent/agent-mode.d.ts

@@ -41,17 +41,33 @@ export declare function nextMode(current: AgentMode): AgentMode;
  * PowerShell, a foreign process resetting the mode), so a rung that depends on
  * NOTHING has to exist underneath it.
  *
- * `ctrl+p` IS MEASURED, not guessed. On the lossy path it arrives as `0x10`,
+ * `ctrl+n` IS MEASURED, not guessed. On the lossy path it arrives as `0x0e`,
  * because Windows generates a real control character for `Ctrl+<letter>` and
  * libuv's character branch forwards it untouched; `Alt+M` produces no bytes at
  * all and `Ctrl+Tab` collapses exactly like `Shift+Tab` (analysis section 3.6).
  * Any idle `Ctrl+<letter>` in the `0x01-0x1A` range would work; this one is free
- * throughout the package - `keymap.ts` claims a/e/w/u/k, `App` claims c/l/t/o,
+ * throughout the package - `keymap.ts` claims a/e/w/u/k, `App` claims c/g/k/l/o/r/t,
  * and `PromptInput` returns early on every `key.ctrl` - so binding it takes no
  * behaviour away from anyone. `Ctrl+B` survives the same path but is tmux's
  * default prefix, and this binding is unconditional and cross-platform, so it
  * would be silently dead for a whole class of macOS and Linux users.
  *
+ * THIS KEY MOVED FROM `ctrl+p`, WHICH NOW OPENS THE PROMPT LIBRARY
+ * (quick-prompt-library D-1). The requirement names `ctrl+p` as the first entry
+ * to a feature users will build muscle memory for, and a key that is sometimes
+ * the prompt library and sometimes the mode switch is a key nobody can learn.
+ * `ctrl+n` was the only idle C0 byte left that survives the same lossy path, and
+ * it carries no conflicting behaviour: readline's `next-history` is `Up`/`Down`
+ * in this product.
+ *
+ * `^N` IS NOT AN ACCEPTABLE SPELLING OF IT, ANYWHERE. `^N` is already this
+ * product's word for TRANSCRIPT LINE OFFSET - the header's scroll chip reads
+ * `^<n>` and the help overlay has a `^N` row about the distance from the bottom -
+ * while every other `^X` in the same row is a key. Writing the mode fallback as
+ * `^N` would put two meanings of `^` on one line, so the header, the help
+ * overlay, `doctor`, the README and the one-shot migration notice all spell out
+ * `Ctrl+N`.
+ *
  * BOUND UNCONDITIONALLY, on every platform. A binding that only exists where it
  * is needed cannot be documented in one place, cannot be taught by muscle
  * memory, and is missing on precisely the machine whose capability probe was
@@ -59,13 +75,13 @@ export declare function nextMode(current: AgentMode): AgentMode;
  *
  * THESE ARE THE LABELS, NOT THE BINDING. Editing a value here changes only what
  * the user is TOLD to press - the hint row, the startup notice, `doctor`. The
- * key that is actually listened for is `input === 'p'` in `App`'s `useInput`,
+ * key that is actually listened for is `input === 'n'` in `App`'s `useInput`,
  * and the help overlay and README spell both keys out in their own title case.
  * Change one without the others and the product advertises a key it ignores.
  */
 export declare const MODE_TOGGLE_KEYS: {
     readonly primary: "shift+tab";
-    readonly fallback: "ctrl+p";
+    readonly fallback: "ctrl+n";
 };
 /** The two tools that only exist because a human is attached. */
 export declare const PLAN_TOOL_NAMES: readonly ["ask_user", "submit_plan"];

package/dist/agent/agent-mode.js

@@ -43,17 +43,33 @@ export function nextMode(current) {
  * PowerShell, a foreign process resetting the mode), so a rung that depends on
  * NOTHING has to exist underneath it.
  *
- * `ctrl+p` IS MEASURED, not guessed. On the lossy path it arrives as `0x10`,
+ * `ctrl+n` IS MEASURED, not guessed. On the lossy path it arrives as `0x0e`,
  * because Windows generates a real control character for `Ctrl+<letter>` and
  * libuv's character branch forwards it untouched; `Alt+M` produces no bytes at
  * all and `Ctrl+Tab` collapses exactly like `Shift+Tab` (analysis section 3.6).
  * Any idle `Ctrl+<letter>` in the `0x01-0x1A` range would work; this one is free
- * throughout the package - `keymap.ts` claims a/e/w/u/k, `App` claims c/l/t/o,
+ * throughout the package - `keymap.ts` claims a/e/w/u/k, `App` claims c/g/k/l/o/r/t,
  * and `PromptInput` returns early on every `key.ctrl` - so binding it takes no
  * behaviour away from anyone. `Ctrl+B` survives the same path but is tmux's
  * default prefix, and this binding is unconditional and cross-platform, so it
  * would be silently dead for a whole class of macOS and Linux users.
  *
+ * THIS KEY MOVED FROM `ctrl+p`, WHICH NOW OPENS THE PROMPT LIBRARY
+ * (quick-prompt-library D-1). The requirement names `ctrl+p` as the first entry
+ * to a feature users will build muscle memory for, and a key that is sometimes
+ * the prompt library and sometimes the mode switch is a key nobody can learn.
+ * `ctrl+n` was the only idle C0 byte left that survives the same lossy path, and
+ * it carries no conflicting behaviour: readline's `next-history` is `Up`/`Down`
+ * in this product.
+ *
+ * `^N` IS NOT AN ACCEPTABLE SPELLING OF IT, ANYWHERE. `^N` is already this
+ * product's word for TRANSCRIPT LINE OFFSET - the header's scroll chip reads
+ * `^<n>` and the help overlay has a `^N` row about the distance from the bottom -
+ * while every other `^X` in the same row is a key. Writing the mode fallback as
+ * `^N` would put two meanings of `^` on one line, so the header, the help
+ * overlay, `doctor`, the README and the one-shot migration notice all spell out
+ * `Ctrl+N`.
+ *
  * BOUND UNCONDITIONALLY, on every platform. A binding that only exists where it
  * is needed cannot be documented in one place, cannot be taught by muscle
  * memory, and is missing on precisely the machine whose capability probe was
@@ -61,11 +77,11 @@ export function nextMode(current) {
  *
  * THESE ARE THE LABELS, NOT THE BINDING. Editing a value here changes only what
  * the user is TOLD to press - the hint row, the startup notice, `doctor`. The
- * key that is actually listened for is `input === 'p'` in `App`'s `useInput`,
+ * key that is actually listened for is `input === 'n'` in `App`'s `useInput`,
  * and the help overlay and README spell both keys out in their own title case.
  * Change one without the others and the product advertises a key it ignores.
  */
-export const MODE_TOGGLE_KEYS = { primary: 'shift+tab', fallback: 'ctrl+p' };
+export const MODE_TOGGLE_KEYS = { primary: 'shift+tab', fallback: 'ctrl+n' };
 /** The two tools that only exist because a human is attached. */
 export const PLAN_TOOL_NAMES = ['ask_user', 'submit_plan'];
 /**

package/dist/agent/controller.d.ts

@@ -1,4 +1,6 @@
 import { type ModelSettingsSnapshot } from './model-profile-settings.js';
+import { type ActiveKit } from '../config/kit-resolution.js';
+import type { KitApplyResult } from '../config/model-kits.js';
 import type { ModelSettingsDraft, ModelSettingsPatch, ModelSettingsSaveResult } from '../config/model-profile-store.js';
 import { type CompactionIdentity } from '../compaction/memory-identity.js';
 /**
@@ -28,7 +30,7 @@ import { type FastStatus } from '../fast/wiring.js';
 import type { FastEventListener, FastSnapshot } from '../fast/types.js';
 import { ContextMeter, type ContextUsageListener } from '../compaction/meter.js';
 import type { CompactionEventListener, CompactionSnapshot, ContextUsageSnapshot } from '../compaction/types.js';
-import type { TodoEventListener, TodoSnapshot } from '../todo/types.js';
+import type { AcceptanceItem, TodoEventListener, TodoSnapshot } from '../todo/types.js';
 import { type HumanInputGate } from '../tools/human-input.js';
 import { SkillService, type ApprovalGate } from '../skills/service.js';
 import { type ConversationResetOptions, type ConversationResetResult } from './conversation-reset.js';
@@ -585,6 +587,37 @@ export declare class AgentController {
      */
     setTodoConfig(patch: Partial<TodoConfig>): TodoConfig;
     getTodoSnapshot(): TodoSnapshot | null;
+    /**
+     * Whether the acceptance table is live for this session: the config key AND the
+     * session switch. Two flags, exactly as `isTodoEnabled` needs `isTodoRegistered`
+     * first - a capability that is on while `todo_write` refuses is not a
+     * capability.
+     */
+    isTodoAcceptanceEnabled(): boolean;
+    /**
+     * Flip the acceptance capability and RE-RENDER THE PROMPT, byte for byte the
+     * contract `setTodoEnabled` records above: the paragraph is part of
+     * `<todo_planning>`, and the tool reads the same live value through its
+     * `hasAcceptance` closure (D-11).
+     */
+    setTodoAcceptanceEnabled(enabled: boolean): void;
+    /**
+     * The acceptance table as DATA, independent of whether it is being drawn.
+     *
+     * The persistence surfaces must use this and not `getTodoSnapshot()?.acceptance`:
+     * the projection is `null` whenever the capability is off, and a checkpoint or a
+     * `/save` that silently dropped the user's verdicts because a display switch was
+     * off would be a data loss dressed as a setting.
+     */
+    getAcceptanceItems(): AcceptanceItem[];
+    /**
+     * `/todo acceptance reset`: every verdict back to `pending`, evidence dropped.
+     *
+     * The count comes BACK from the store rather than being recomputed by the
+     * command, so the notice and the state cannot disagree about how many criteria
+     * moved.
+     */
+    reopenAcceptance(): number;
     /**
      * Subscribe to the CLI-local todo event stream.
      *
@@ -592,8 +625,12 @@ export declare class AgentController {
      * every caller can subscribe unconditionally.
      */
     subscribeTodos(listener: TodoEventListener): () => void;
-    /** `/resume`. An absent or empty list CLEARS — see `TodoStore.restore` (P0-2). */
-    restoreTodos(items: unknown): void;
+    /**
+     * `/resume`. An absent or empty todo list CLEARS both tables — see
+     * `TodoStore.restore` (P0-2). `acceptance` is optional so every existing caller
+     * keeps compiling, and an omitted table reads as "this file has none".
+     */
+    restoreTodos(items: unknown, acceptance?: unknown): void;
     /**
      * `/todo clear` and `/clear screen`. A user override of a projection (I-2's
      * standing exception). `/todo` refuses mid-run before it gets here;
@@ -629,6 +666,33 @@ export declare class AgentController {
     saveModelSettingsPatch(patch: ModelSettingsPatch): ModelSettingsSaveResult;
     /** Refresh both role connections and credentials while retaining conversation state. */
     reloadModelSettings(): ModelSettingsSaveResult;
+    /**
+     * The narrow port `agent/kit-apply.ts` runs against (agent-model-kits §6.3).
+     *
+     * A LITERAL ADAPTER RATHER THAN `this`: the port's member names (`isBusy`,
+     * `hasKey`, `preview`) are deliberately NOT the controller's names, because a
+     * port that shadows a public method is one signature change away from silently
+     * calling itself. Three of the five members are one-liners onto existing
+     * methods; nothing here is new state.
+     */
+    private kitApplyPort;
+    /**
+     * Apply a kit for THIS SESSION ONLY (D-7 / I-5): no commit, ever.
+     *
+     * Failed applies leave the live config and the file untouched (I-10), because
+     * the whole operation is one prepare-and-apply with no commit in between.
+     */
+    applySessionKit(kitId: string): KitApplyResult;
+    /** Release the session kit and return to the config file's bindings (RV-P1-6). */
+    clearSessionKit(): KitApplyResult;
+    /**
+     * Which kit is in effect right now (session > default > disk bindings).
+     *
+     * DERIVED, never stored (D-5): the header chip, the settings root row and
+     * `/kit` all read this, so there is exactly one answer to "which preset am I
+     * running" and no copy of it that can drift.
+     */
+    getActiveKit(): ActiveKit;
     /** Prebuild all potentially failing model/prompt work before committing the file. */
     prepareModelSettingsSnapshot(config: CliConfig): ModelSettingsSnapshot;
     /** Apply prepared state synchronously, without rebuilding the controller or its history. */