lcode-cli 0.32.0__tar.gz → 0.33.0__tar.gz

{lcode_cli-0.32.0 → lcode_cli-0.33.0}/CHANGELOG.md

@@ -6,6 +6,23 @@ All notable changes to lcode are documented here. The format follows
 
 ## [Unreleased]
 
+## [0.33.0] - 2026-10-14
+
+### Added
+
+- Steer while it works. Type while a request runs and press Enter: the line goes to the model at
+  its next step, as a note from you, without interrupting anything. The spinner shows what you're
+  typing and the notes waiting. Esc pauses after the current step, so you can look, then go on with
+  or without a note. Notes count as your requests for summaries and the review, and anything that
+  never reached the model goes back into the prompt. A command that runs for a while now shows a
+  spinner with its elapsed time. See
+  [Steer while it works](https://nasser1941.github.io/lcode/usage/#steer-while-it-works).
+
+### Fixed
+
+- "Where you left off" (`lcode -c`) and session titles showed lcode's own messages to the model,
+  such as a failed check, as if you had written them. They show your last request now.
+
 ## [0.32.0] - 2026-10-14
 
 ### Added
@@ -691,7 +708,8 @@ First public release.
 - `AGENTS.md` project instructions and `/init` to generate them.
 - One-line installer for Linux and macOS.
 
-[Unreleased]: https://github.com/nasser1941/lcode/compare/v0.32.0...HEAD
+[Unreleased]: https://github.com/nasser1941/lcode/compare/v0.33.0...HEAD
+[0.33.0]: https://github.com/nasser1941/lcode/compare/v0.32.0...v0.33.0
 [0.32.0]: https://github.com/nasser1941/lcode/compare/v0.31.0...v0.32.0
 [0.31.0]: https://github.com/nasser1941/lcode/compare/v0.30.0...v0.31.0
 [0.30.0]: https://github.com/nasser1941/lcode/compare/v0.29.0...v0.30.0

{lcode_cli-0.32.0 → lcode_cli-0.33.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.5
 Name: lcode-cli
-Version: 0.32.0
+Version: 0.33.0
 Summary: A local-first terminal coding agent powered by open-weight models on your own GPU or Mac, via Ollama.
 Project-URL: Homepage, https://nasser1941.github.io/lcode/
 Project-URL: Documentation, https://nasser1941.github.io/lcode/

{lcode_cli-0.32.0 → lcode_cli-0.33.0}/docs/quickstart.md

@@ -12,7 +12,7 @@ lcode
 
 ```text
 ╭──────────────────────────────────────────────╮
-│ lcode v0.32.0 — local coding agent           │
+│ lcode v0.33.0 — local coding agent           │
 │                                              │
 │ model    lcode-qwen3.6-35b                   │
 │ context  256K tokens                         │

{lcode_cli-0.32.0 → lcode_cli-0.33.0}/docs/usage.md

@@ -44,6 +44,8 @@ lcode update [--check]       # update lcode, and Ollama when it runs in Docker (
 | ++enter++ | Send |
 | ++esc++ ++enter++ or a trailing `\` | New line |
 | ++ctrl+c++ | Interrupt the model or a running command |
+| Typing while it works | A note for the model, read at its next step ([Steer while it works](#steer-while-it-works)) |
+| ++esc++ while it works | Pause after the current step |
 | ++ctrl+d++ | Quit |
 | ++shift+tab++ | Cycle permission mode: ask → plan → auto-edit → yolo |
 | ++tab++ | Complete slash commands and `@` paths |
@@ -231,9 +233,10 @@ Much of the waiting with a local model isn't the model thinking: it's loading th
 a long conversation before the first word. lcode says which, in the status line under a request:
 
 ```text
-⠋ Loading qwen3.6-35b into memory… (0 tokens, 9s · Ctrl+C to stop)
-⠋ Reading the conversation (18K tokens)… (0 tokens, 21s · Ctrl+C to stop)
-⠋ Thinking… (312 tokens, 27s · Ctrl+C to stop)
+⠋ Loading qwen3.6-35b into memory… (0 tokens, 9s · type a note · Esc pauses · Ctrl+C stops)
+⠋ Reading the conversation (18K tokens)… (0 tokens, 21s · type a note · Esc pauses · Ctrl+C stops)
+⠋ Thinking… (312 tokens, 27s · type a note · Esc pauses · Ctrl+C stops)
+⠋ Running the command… (48s · type a note · Esc pauses · Ctrl+C stops)
 ```
 
 It prints a line only when something worth knowing happens, and nothing while the model simply
@@ -252,6 +255,42 @@ While the model loads in the background at startup, the status bar says
 part of them in system RAM. Dense ones slow down a lot, to a few tokens a second, which is why they
 get the warning.
 
+## Steer while it works
+
+A request can take minutes with a local model. You don't have to stop it to change course: type
+while it works, and press ++enter++. The line goes to the model at its next step, after the step it's
+on finishes, as a note from you. Nothing is interrupted or thrown away.
+
+```text
+● edit_file(tally.py)
+  ✓ Edited tally.py
+✎ Note for the model: also add a --min-count N option that skips rare words
+```
+
+What you type shows under the spinner, since the terminal doesn't echo it while lcode works:
+
+```text
+⠋ Thinking… (212 tokens, 14s · type a note · Esc pauses · Ctrl+C stops)
+  ✎ also add a --min-count N opt▏  Enter adds it as a note
+```
+
+- A note typed while the model writes its final answer still counts: the model reads it before the
+  request ends.
+- Press ++esc++ to pause after the current step, to look at what it did. Then press ++enter++ to go
+  on, type a note to go on with it, or press ++ctrl+c++ to stop the request. Press ++esc++ again
+  before the step ends if you change your mind.
+- Notes count as your requests: they're kept when old messages are
+  [summarized](how-it-works.md#context-management), and the [review](#a-review-before-the-answer)
+  checks the change against them too.
+- When lcode asks you something, such as a permission, the keyboard is yours again for the answer.
+  Keys you typed just before the question appeared are dropped, so a note you were finishing can't
+  answer it by accident.
+- What never reached the model, such as a note typed as the request ended or after ++ctrl+c++, is
+  put back in the prompt.
+
+Steering works in the interactive session on macOS and Linux. `lcode -p` and the editor
+integration don't read the keyboard during a request.
+
 ## Notifications
 
 Local models take a while, so you'll often switch to something else during a request. lcode shows

{lcode_cli-0.32.0 → lcode_cli-0.33.0}/src/lcode/__init__.py

@@ -3,7 +3,7 @@
 From a program: `from lcode import Session` (see lcode.api).
 """
 
-__version__ = "0.32.0"
+__version__ = "0.33.0"
 
 API = ("Session", "Result", "Options", "SetupError")
 

{lcode_cli-0.32.0 → lcode_cli-0.33.0}/src/lcode/agent.py

@@ -52,6 +52,7 @@ from lcode.permissions import Permissions
 from lcode.render import MarkdownStreamer
 from lcode.repeats import Repeats
 from lcode.sandbox import Sandbox, SandboxError, project_root
+from lcode.steer import NOTE_INTRO, Steering
 from lcode.tools import (
     LSP_SCHEMA,
     MEMORY_SCHEMA,
@@ -224,10 +225,12 @@ class Progress:
     arrives, for example while the model writes a long file into a tool call.
     """
 
-    def __init__(self, phase: str = "Thinking") -> None:
+    def __init__(self, phase: str = "Thinking", steering: Steering | None = None, tokens: bool = True) -> None:
         self.started = self.last_chunk = time.time()
         self.phase = phase
         self.chars = 0
+        self.steering = steering  # the note being typed and the notes waiting, shown on a second line
+        self.tokens = tokens  # False for a command running: no tokens to count
 
     def update(self, phase: str, text: str) -> None:
         self.phase, self.chars, self.last_chunk = phase, self.chars + len(text), time.time()
@@ -240,9 +243,11 @@ class Progress:
         now = time.time()
         streaming = self.phase in ("Thinking", "Writing")  # loading and reading say what they are throughout
         phase = self.phase if now - self.last_chunk < 3 or not streaming else "Working"
-        return Text.from_markup(
-            f"[cyan]{phase}…[/] [dim]({self.chars // 4} tokens, {now - self.started:.0f}s · Ctrl+C to stop)[/]"
-        )
+        tokens = f"{self.chars // 4} tokens, " if self.tokens else ""
+        keys = "Ctrl+C to stop" if self.steering is None else "type a note · Esc pauses · Ctrl+C stops"
+        line = Text.from_markup(f"[cyan]{phase}…[/] [dim]({tokens}{now - self.started:.0f}s · {keys})[/]")
+        below = self.steering.status() if self.steering is not None else None
+        return line if below is None else Text("\n").join([line, below])
 
 
 @dataclass
@@ -321,6 +326,9 @@ class Agent:
         self.read_only = False  # only read-only shell commands
         self.no_changes = ""  # set during a review: why nothing may change (read-only tools only)
         self.cancel: threading.Event | None = None  # set from another thread to stop
+        self.steering: Steering | None = None  # notes typed and pauses asked for during a request (lcode.repl)
+        self.turn_requests: list[str] = []  # the request, then the notes added to it while it ran
+        self.unsent = ""  # typed during the last request but never sent to the model: back in the prompt (lcode.repl)
         self.on_tool = None  # called with (name, arguments) before each tool runs
         self.jobs = Jobs()  # background commands (shared with subagents)
         self.turn_started = 0.0  # when the running request started (time.monotonic)
@@ -749,7 +757,7 @@ class Agent:
         """Stream one model response, rendering thinking/content live. Returns the assistant message."""
         content, thinking, tool_calls, final = "", "", [], {}
         md = MarkdownStreamer(self.console)
-        progress = Progress()
+        progress = Progress(steering=self.steering)
         watch = self.before_request(progress)
         status = None
         printed_thinking = False
@@ -845,15 +853,10 @@ class Agent:
         diff, changed = review.diff_of(self.tools.before, self.current_text, budget)
         if changed < review.MIN_CHANGED_LINES:
             return []
-        request = next(
-            (
-                context_tools.request_text(str(m.get("content", "")))
-                for m in reversed(self.messages)
-                if m.get("role") == "user"
-                and not str(m.get("content", "")).startswith(("[lcode]", sessions.SUMMARY_PREFIX))
-            ),
-            "",
-        )
+        request = self.turn_requests[0] if self.turn_requests else ""
+        if len(self.turn_requests) > 1:
+            added = "\n".join(f"- {note}" for note in self.turn_requests[1:])
+            request += f"\n\nWhat the user added while you worked:\n{added}"
         try:
             with self.console.status("Reviewing the change against your request…", spinner="dots"):
                 problems = review.run(self, request, answer, diff)
@@ -1069,6 +1072,7 @@ class Agent:
         self.turn_status = "running"
         note = planning.NOTE if self.planning() and self.allowed_tools is None else ""
         self.messages.append({"role": "user", "content": self.expand_mentions(user_text) + note})
+        self.turn_requests = [user_text.strip()]
         self.repeats.reset()
         self.check_results = []
         self.review_results = None
@@ -1081,6 +1085,8 @@ class Agent:
         for step in range(self.max_steps):
             if self.cancel is not None and self.cancel.is_set():
                 raise KeyboardInterrupt
+            if step and self.steering is not None:
+                self.steer()
             if step and self.autosave:
                 self.save()  # a crash or a closed terminal loses at most this step
             self.maybe_compact()
@@ -1126,6 +1132,8 @@ class Agent:
                 )
                 continue
             if not calls:
+                if self.steering is not None and self.steering.waiting():
+                    continue  # the user added a note while the model wrote its answer: it reads the note first
                 check = self.project_check() if changed and not verified else None
                 if check is not None and check_rounds < checks.MAX_ROUNDS:
                     check_rounds += 1
@@ -1241,6 +1249,34 @@ class Agent:
         self.turn_status = "success"
         self.print_stats()
 
+    def steer(self) -> None:
+        """Between two steps: pause if the user pressed Esc, then hand the model the notes typed meanwhile."""
+        steering = self.steering
+        if steering is None:
+            return
+        if steering.take_pause():
+            self.console.print(
+                Text(
+                    "⏸ Paused. Press Enter to go on, type a note to go on with it, or press Ctrl+C to stop the request.",
+                    style="yellow",
+                )
+            )
+            note = steering.ask("  ⏸ ").strip()
+            if note:
+                steering.add(note)
+        notes = steering.take()
+        if not notes:
+            return
+        self.turn_requests += notes
+        self.messages.append({"role": "user", "content": "\n\n".join(notes) + "\n\n" + NOTE_INTRO})
+        for note in notes:
+            shown = note if len(note) <= 200 else note[:200] + "…"
+            self.console.print(Text(f"✎ Note for the model: {shown}", style="cyan"))
+
+    def activity(self, phase: str):
+        """A spinner with the elapsed time while something runs (a command), with the note being typed under it."""
+        return self.console.status(Progress(phase, self.steering, tokens=False), spinner="dots")
+
     def print_stats(self) -> None:
         pct = 100 * self.ctx_used / self.settings.context
         self.console.print(

{lcode_cli-0.32.0 → lcode_cli-0.33.0}/src/lcode/permissions.py

@@ -198,6 +198,7 @@ class Permissions:
         self.always: set[str] = set()
         self.rules = Rules()  # allow and deny lists from the settings (lcode.hooks)
         self.on_prompt = None  # called with the title before lcode asks the user (notification hooks)
+        self.read_line: Callable[[str], str] | None = None  # asks instead of input() (lcode.steer, during a request)
         # Decides instead of asking (lcode.api, lcode acp): True, False, or a message saying why not.
         self.approve: Callable[[dict], bool | str] | None = None
 
@@ -243,7 +244,8 @@ class Permissions:
         if kind == "mcp":
             scope = scope.replace(":", " › ", 1)
         try:
-            answer = input(f"  Allow? [y]es / [a]lways for '{scope}' this session / [n]o (+ optional reason): ")
+            question = f"  Allow? [y]es / [a]lways for '{scope}' this session / [n]o (+ optional reason): "
+            answer = (self.read_line or input)(question)
         except EOFError:
             return False, "The user denied this action (no input available)."
         answer = answer.strip()

{lcode_cli-0.32.0 → lcode_cli-0.33.0}/src/lcode/repl.py

@@ -36,6 +36,7 @@ from lcode.checkpoints import Checkpoint, CheckpointError, Restore
 from lcode.config import PERMISSION_MODES, STATE_DIR, ConfigError, format_tokens, parse_context
 from lcode.hardware import Hardware
 from lcode.ollama import OllamaError
+from lcode.steer import KeyReader, Steering
 from lcode.tools import IGNORE_DIRS
 
 COMMANDS = {
@@ -874,6 +875,7 @@ def git_command(agent: Agent, cmd: str, arg: str) -> None:
 
 def run_safely(agent: Agent, text: str) -> None:
     started = time.monotonic()
+    reader = start_steering(agent) if agent.interactive else None
     try:
         agent.run_turn(text)
     except KeyboardInterrupt:
@@ -881,10 +883,33 @@ def run_safely(agent: Agent, text: str) -> None:
     except OllamaError as e:
         agent.console.print(f"[red]{e}[/]")
     finally:
+        if reader is not None:
+            stop_steering(agent, reader)
         agent.save()
         agent.finished(text, time.monotonic() - started)
 
 
+def start_steering(agent: Agent) -> KeyReader | None:
+    """Read the keyboard while the request runs: a typed line becomes a note for the model, Esc pauses."""
+    steering = Steering()
+    reader = KeyReader(steering)
+    if not reader.start():
+        return None  # not a terminal lcode can read keys from (or Windows): the request runs as before
+    agent.steering = steering
+    agent.perms.read_line = steering.ask  # questions take the keyboard back meanwhile
+    return reader
+
+
+def stop_steering(agent: Agent, reader: KeyReader) -> None:
+    reader.stop()
+    agent.steering = None
+    agent.perms.read_line = None
+    unsent = reader.steering.leftover()
+    if unsent:
+        agent.unsent = unsent
+        agent.console.print(Text("✎ What you typed didn't reach the model; it's in the prompt below.", style="cyan"))
+
+
 def session_start(agent: Agent) -> None:
     """session_start hooks: their output is shown, and with feedback = true also given to the model."""
     for outcome in agent.hooks.run("session_start", {"cwd": str(agent.cwd), "session": agent.session_id}, agent.cwd):
@@ -1115,13 +1140,13 @@ def resume_session(agent: Agent, info: sessions.SessionInfo) -> None:
 
 def print_recap(agent: Agent) -> None:
     """Show the last request and the start of the last answer, so it's clear where things left off."""
-    last_user = next((m for m in reversed(agent.messages) if m.get("role") == "user"), None)
+    requests = context_tools.user_requests(agent.messages)  # the user's own words, not lcode's notes to the model
     last_answer = next(
         (m for m in reversed(agent.messages) if m.get("role") == "assistant" and m.get("content", "").strip()), None
     )
-    if not last_user:
+    if not requests:
         return
-    lines = [f"[bold]You:[/] {escape(sessions.title_from([last_user]))}"]
+    lines = [f"[bold]You:[/] {escape(sessions.title_from([{'role': 'user', 'content': requests[-1]}]))}"]
     if last_answer:
         answer = " ".join(last_answer["content"].split())
         answer = answer if len(answer) <= 300 else answer[:299].rstrip() + "…"
@@ -1153,7 +1178,8 @@ def repl(agent: Agent, prompt: str | None, hardware: Hardware, cont: bool = Fals
     while True:
         try:
             agent.console.print()
-            line = session.prompt(HTML("<ansicyan><b>❯ </b></ansicyan>")).strip()
+            unsent, agent.unsent = agent.unsent, ""
+            line = session.prompt(HTML("<ansicyan><b>❯ </b></ansicyan>"), default=unsent).strip()
         except KeyboardInterrupt:
             continue
         except EOFError:

{lcode_cli-0.32.0 → lcode_cli-0.33.0}/src/lcode/sessions.py

@@ -45,7 +45,9 @@ def title_from(messages: list[dict]) -> str:
         if content.startswith(SUMMARY_PREFIX):
             compacted = True
             continue
-        text = re.split(r"\n\n<(?:file|directory) path=", content, maxsplit=1)[0]
+        if content.startswith("[lcode]"):
+            continue  # lcode's own message to the model (a failed check, a review), not the user's words
+        text = re.split(r"\n\n(?:<(?:file|directory|image) path=|\[lcode[:\]] )", content, maxsplit=1)[0]
         text = " ".join(text.split())
         if text:
             return text if len(text) <= TITLE_LENGTH else text[: TITLE_LENGTH - 1].rstrip() + "…"

lcode_cli-0.33.0/src/lcode/steer.py

@@ -0,0 +1,287 @@
+"""Steer a request while it runs: type a note for the model, or press Esc to pause after the step.
+
+With a local model a request can take minutes. Ctrl+C throws away the step in progress; a note doesn't
+interrupt anything: a line typed while lcode works is kept and handed to the model before its next step, and the
+spinner line shows what's being typed and what's waiting. Esc asks lcode to stop after the current step, so you
+can look at what it did and then go on, with or without a note.
+
+`Steering` holds the state and turns keystrokes into notes; it knows nothing about terminals, so it's easy to
+test. `KeyReader` reads the keys from the terminal in the background, without echoing them (that would garble
+the output), and hands the keyboard back whenever lcode asks a question (`Steering.ask`).
+"""
+
+from __future__ import annotations
+
+import codecs
+import contextlib
+import os
+import select
+import sys
+import threading
+import time
+from collections.abc import Callable, Iterator
+
+from rich.text import Text
+
+try:
+    import termios
+except ImportError:  # Windows: no steering, everything else works
+    termios = None  # type: ignore[assignment]
+
+MAX_NOTE = 4000  # characters in one note
+SHOWN = 45  # characters of the line being typed shown under the spinner (the line fits 80 columns)
+ESC_WAIT = 0.03  # an Esc followed by more bytes this soon is an arrow or function key, not the Esc key
+TOO_QUICK = 0.4  # an Enter this soon after a question appears was meant for the note being typed
+NOTE_INTRO = (
+    "[lcode] The user wrote this while you were working. Take it into account from here on; it doesn't mean "
+    "starting over unless it says so."
+)
+
+
+class Steering:
+    """Notes typed during a request, the line being typed, and whether to pause after the step."""
+
+    def __init__(self) -> None:
+        self.lock = threading.Lock()
+        self.typed = ""  # the line being typed
+        self.notes: list[str] = []  # finished lines, waiting for the model's next step
+        self.pause = False  # stop after the current step
+        self.reader: KeyReader | None = None
+        self.read_line: Callable[[str], str] | None = None  # input() when None (looked up when asking)
+
+    # Input ------------------------------------------------------------------------------------------------
+
+    def feed(self, keys: str) -> None:
+        """Keystrokes as text: printable characters, Enter, Backspace, Ctrl+U/Ctrl+W, Esc and escape sequences.
+
+        A lone "\\x1b" is the Esc key (KeyReader waits a moment for the rest of a sequence before it hands one
+        over). Everything else that starts with it is skipped: the sequences arrow and function keys send,
+        Alt+key, and a terminal's answers to queries (cursor position, colours).
+        """
+        with self.lock:
+            i = 0
+            while i < len(keys):
+                ch = keys[i]
+                i += 1
+                if ch == "\x1b":
+                    after = keys[i] if i < len(keys) else ""
+                    if after == "[":  # CSI: parameters, then a final byte in @..~
+                        i += 1
+                        while i < len(keys) and not ("@" <= keys[i] <= "~"):
+                            i += 1
+                        i += 1
+                    elif after == "O":  # SS3: one more character (F1-F4, keypad)
+                        i += 2
+                    elif after in ("]", "P", "_", "^"):  # a string (OSC, DCS…), ended by BEL or ESC \
+                        end = min(
+                            (j for j in (keys.find("\x07", i), keys.find("\x1b\\", i)) if j >= 0), default=len(keys)
+                        )
+                        i = end + (2 if keys.startswith("\x1b\\", end) else 1)
+                    elif after in ("", "\x1b"):  # the Esc key (twice in a row: two presses)
+                        self.pause = not self.pause
+                    else:
+                        i += 1  # Alt+key
+                elif ch in "\r\n":
+                    note = self.typed.strip()
+                    self.typed = ""
+                    if note:
+                        self.notes.append(note[:MAX_NOTE])
+                elif ch in "\x7f\x08":
+                    self.typed = self.typed[:-1]
+                elif ch == "\x15":  # Ctrl+U: clear the line
+                    self.typed = ""
+                elif ch == "\x17":  # Ctrl+W: delete the last word
+                    self.typed = self.typed.rstrip()
+                    self.typed = self.typed[: self.typed.rfind(" ") + 1] if " " in self.typed else ""
+                elif ch == "\t":
+                    self.typed += " "
+                elif ch >= " " and len(self.typed) < MAX_NOTE:
+                    self.typed += ch
+
+    # For the agent ----------------------------------------------------------------------------------------
+
+    def add(self, note: str) -> None:
+        with self.lock:
+            self.notes.append(note[:MAX_NOTE])
+
+    def waiting(self) -> bool:
+        with self.lock:
+            return bool(self.notes)
+
+    def take(self) -> list[str]:
+        """The notes waiting for the model (and forget them)."""
+        with self.lock:
+            notes, self.notes = self.notes, []
+            return notes
+
+    def take_pause(self) -> bool:
+        """Whether a pause was asked for (and clear the request)."""
+        with self.lock:
+            asked, self.pause = self.pause, False
+            return asked
+
+    def leftover(self) -> str:
+        """What never reached the model: notes still waiting and the unfinished line, for the next prompt."""
+        with self.lock:
+            text = "\n".join([*self.notes, self.typed.strip()]).strip()
+            self.notes, self.typed = [], ""
+            return text
+
+    def ask(self, question: str) -> str:
+        """Ask the user a question while the keyboard is being read for notes (permission prompts, the pause).
+
+        Keys typed before the question appeared are dropped, and an Enter pressed right as it appeared asks
+        once more: it was meant to finish a note, not to answer (an empty answer often means yes).
+        """
+        hold = self.reader.suspended() if self.reader is not None else contextlib.nullcontext()
+        with hold:
+            asked = time.monotonic()
+            answer = (self.read_line or input)(question)
+            if not answer.strip() and time.monotonic() - asked < TOO_QUICK:
+                answer = (self.read_line or input)(question)
+            return answer
+
+    def status(self) -> Text | None:
+        """The line under the spinner: what's being typed, notes waiting, a pause coming."""
+        with self.lock:
+            typed, waiting, pause = self.typed, len(self.notes), self.pause
+        parts: list[Text] = []
+        if typed:
+            shown = typed if len(typed) <= SHOWN else "…" + typed[-SHOWN:]
+            parts.append(Text.assemble(("✎ ", "cyan"), shown, ("▏", "cyan"), ("  Enter adds it as a note", "dim")))
+        if waiting:
+            what = "a note" if waiting == 1 else f"{waiting} notes"
+            parts.append(Text(f"✎ {what} waiting for the model's next step", style="cyan"))
+        if pause:
+            parts.append(Text("⏸ Pausing after this step (Esc again to go on without pausing)", style="yellow"))
+        if not parts:
+            return None
+        return Text("\n").join(parts)
+
+
+class KeyReader:
+    """Reads the keyboard in the background while a request runs, and gives the keys to a Steering.
+
+    The terminal stays in "cbreak" mode meanwhile: keys arrive one by one and aren't echoed, while Ctrl+C still
+    interrupts. `suspended()` hands the keyboard back, in its usual mode, for a question.
+    """
+
+    def __init__(self, steering: Steering, fd: int | None = None) -> None:
+        self.steering = steering
+        self.fd = fd  # stdin when None
+        self.saved: list | None = None
+        self.thread: threading.Thread | None = None
+        self.stopping = threading.Event()
+        self.hold = threading.Event()  # asked to stop reading for a while
+        self.held = threading.Event()  # ...and the thread has stopped
+        self.resume = threading.Event()
+        self.decoder = codecs.getincrementaldecoder("utf-8")(errors="replace")
+
+    def start(self) -> bool:
+        """Start reading; False when this isn't a terminal lcode can read keys from."""
+        if termios is None:
+            return False
+        try:
+            if self.fd is None:
+                self.fd = sys.stdin.fileno()
+            if not os.isatty(self.fd):
+                return False
+        except (OSError, ValueError):  # no real stdin (tests, some IDE consoles)
+            return False
+        try:
+            self.saved = termios.tcgetattr(self.fd)
+            self._cbreak()
+        except termios.error:
+            self.saved = None
+            return False
+        self.steering.reader = self
+        self.thread = threading.Thread(target=self._run, name="lcode-keys", daemon=True)
+        self.thread.start()
+        return True
+
+    def stop(self) -> None:
+        self.stopping.set()
+        self.resume.set()
+        if self.thread is not None:
+            self.thread.join(timeout=1)
+        self._restore()
+        self.steering.reader = None
+
+    def __enter__(self) -> KeyReader:
+        self.start()
+        return self
+
+    def __exit__(self, *exc) -> None:
+        self.stop()
+
+    @contextlib.contextmanager
+    def suspended(self) -> Iterator[None]:
+        """The keyboard back in its usual mode (echo, line editing) for a question, with typed-ahead keys dropped."""
+        if self.thread is None or not self.thread.is_alive():
+            yield
+            return
+        self.resume.clear()
+        self.hold.set()
+        self.held.wait(timeout=1)
+        self._restore()
+        with contextlib.suppress(termios.error):
+            termios.tcflush(self.fd, termios.TCIFLUSH)
+        try:
+            yield
+        finally:
+            with contextlib.suppress(termios.error):
+                self._cbreak()
+            self.hold.clear()
+            self.held.clear()
+            self.resume.set()
+
+    def _cbreak(self) -> None:
+        mode = termios.tcgetattr(self.fd)
+        mode[3] &= ~(termios.ECHO | termios.ICANON)  # lflags; ISIG stays, so Ctrl+C still interrupts
+        mode[6][termios.VMIN] = 1
+        mode[6][termios.VTIME] = 0
+        termios.tcsetattr(self.fd, termios.TCSANOW, mode)
+
+    def _restore(self) -> None:
+        if self.saved is not None:
+            with contextlib.suppress(termios.error):
+                termios.tcsetattr(self.fd, termios.TCSANOW, self.saved)
+
+    def _run(self) -> None:
+        while not self.stopping.is_set():
+            if self.hold.is_set():
+                self.held.set()
+                self.resume.wait()
+                continue
+            try:
+                ready, _, _ = select.select([self.fd], [], [], 0.1)
+                if not ready or self.hold.is_set():
+                    continue
+                data = os.read(self.fd, 1024)
+                if not data:
+                    return  # the terminal is gone
+                while incomplete_escape(data):
+                    more, _, _ = select.select([self.fd], [], [], ESC_WAIT)  # the rest of an escape sequence?
+                    if not more:
+                        break  # a lone "\x1b" is then the Esc key
+                    data += os.read(self.fd, 1024)
+            except (OSError, ValueError):
+                return
+            self.steering.feed(self.decoder.decode(data))
+
+
+def incomplete_escape(data: bytes) -> bool:
+    """Whether the bytes end in the middle of an escape sequence (or with a "\\x1b" that may start one)."""
+    start = data.rfind(b"\x1b")
+    if start < 0:
+        return False
+    tail = data[start:]
+    if tail == b"\x1b":
+        return True
+    if tail.startswith(b"\x1b["):
+        return not any(0x40 <= b <= 0x7E for b in tail[2:])
+    if tail.startswith(b"\x1bO"):
+        return len(tail) < 3
+    if tail[1:2] in (b"]", b"P", b"_", b"^"):
+        return b"\x07" not in tail  # (one ended by ESC \ has its last ESC there: complete)
+    return False