Compaction: a button, and automatically when the window fills

A long conversation eventually just stops working. Compaction summarises the
earlier turns and sends the summary in their place.

The messages are kept. They stay in the transcript behind a collapsed
divider and simply stop being part of the request, which is what makes the
button safe to press and automatic compaction safe to have at all: a summary
that came out badly is a bad turn, not a lost conversation.

Stored on the Chat, not as a synthetic Message. A synthetic row needs a
role -- `system` breaks the one-system-message rule the moment build_messages
emits it beside the harness, and user/assistant makes it a turn people can
edit, regenerate from and copy, indistinguishable from a real one in all
four places a bubble is rendered. Worse, "editing rewinds, it does not
branch" would silently delete it and leave no marker that compaction had
happened at all.

The summary goes out as a user turn and an assistant turn, not one. A
leading assistant breaks templates requiring the first non-system message to
be user; a lone leading user produces user, user whenever the kept history
starts on a user turn -- which it always does, because the cutoff lands on a
finished reply.

compacted_through_id is a plain id rather than a foreign key: migrations.py
compiles only the column type, so a REFERENCES clause would exist on a fresh
database and not on an upgraded one, and a constraint half the fleet has is
worse than none. cutoff_message validates it on every read instead, and a
rewind past the boundary clears it.

Compacting again summarises only the delta, with the previous summary
supplied to be subsumed. Re-summarising the whole chat each time grows
quadratically and eventually exceeds the window it is protecting.

Automatically at the top of _run, not in post_message: that route's contract
is to return immediately and leave the slow part to a resumable connection,
and it also means build_request is called once, after compaction, with no
second assembly path. The trigger is the last reply's recorded usage plus an
estimate of the new turn -- retrospective because true prompt_tokens are only
knowable after a response, plus the delta because otherwise fifty thousand
characters pasted into the composer overflow a window that read 90% last
turn. It never fires when the context length is unknown. It does fire on
estimated counts, which is safe here precisely because nothing is lost.

_maybe_compact never raises: a failure logs and sends the uncompacted
request. A `status` event says "Summarising earlier messages…" in the
meantime, because a silent multi-second pause before the first token is what
a hang looks like.

The wording is three fragments under Admin - Prompts. Clearing task.compact
turns compaction off entirely.

Also adds compaction.moment(): SQLite does not store the offset, so a row
loaded from disk is naive while one in the session's identity map keeps its
tzinfo, and comparing the two raises. Every comparison here is between
exactly those.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

src/lembas/api/admin.py

@@ -58,6 +58,7 @@ async def save_general(
     instance_name: str = Form("LLeMbas"),
     allow_signup: bool = Form(False),
     system_prompt: str = Form(""),
+    compact_threshold: int = Form(95),
 ) -> Response:
     """Save instance settings.
 
@@ -70,6 +71,12 @@ async def save_general(
             "instance_name": instance_name.strip()[:120] or "LLeMbas",
             "allow_signup": allow_signup,
             "system_prompt": system_prompt.strip()[:8000],
+            # 0 is "never"; anything else is clamped into a band where it can
+            # do some good. 100 is useless -- you cannot compact after
+            # overflowing -- and below 50 it fires while there is plenty left.
+            "compact_threshold": (
+                0 if compact_threshold <= 0 else min(max(compact_threshold, 50), 99)
+            ),
         },
     )
     log.info("registration %s by %s", "opened" if allow_signup else "closed", user.email)

src/lembas/api/chats.py

@@ -19,9 +19,11 @@ from lembas.db.session import session_scope
 from lembas.security import permissions
 from lembas.services import audio as audio_service
 from lembas.services import chat as chat_service
+from lembas.services import compaction as compaction_service
 from lembas.services import files as files_service
 from lembas.services import generation as generation_service
 from lembas.services import metrics as metrics_service
+from lembas.services import prompts as prompts_service
 from lembas.services import sse
 from lembas.services import tools as tools_service
 from lembas.services.markdown import escape_text, render_markdown
@@ -202,6 +204,64 @@ def _pretty(payload: dict) -> str:
     return text
 
 
+@router.post("/{chat_id}/compact")
+async def compact_chat(request: Request, db: Db, user: RequiredUser, chat_id: str) -> Response:
+    """Summarise the earlier turns and stop sending them.
+
+    No permission of its own: compaction changes only what one chat sends
+    upstream, and gating it would mean answering "why can this user not tidy
+    their own conversation".
+    """
+    chat = _owned_chat(db, chat_id, user.id)
+
+    unfinished = db.scalar(
+        select(Message).where(Message.chat_id == chat.id, Message.complete.is_(False))
+    )
+    if unfinished is not None:
+        # Summarising a transcript that is still being written races
+        # build_request. Queuing it is a state machine nobody asked for.
+        raise HTTPException(
+            status.HTTP_409_CONFLICT, "Wait for the current reply to finish, then compact."
+        )
+
+    template = prompts_service.resolve(db, "task.compact")
+    if not template.strip():
+        raise HTTPException(
+            status.HTTP_409_CONFLICT,
+            "Compaction is turned off: its prompt is empty under Admin → Prompts.",
+        )
+
+    upto = compaction_service.last_complete(db, chat)
+    if upto is None:
+        raise HTTPException(
+            status.HTTP_409_CONFLICT, "There is nothing here to summarise yet."
+        )
+
+    endpoint, model_id = chat_service.resolve_endpoint(db, chat)
+    transcript = compaction_service.transcript(db, chat, upto=upto)
+    previous = compaction_service.previous_summary_block(chat)
+
+    summary = await chat_service.summarise_for_compaction(
+        endpoint,
+        model_id,
+        transcript=transcript,
+        previous_summary=previous,
+        template=template,
+    )
+    if not summary:
+        raise HTTPException(
+            status.HTTP_409_CONFLICT, "The model returned no summary, so nothing changed."
+        )
+
+    compaction_service.apply(chat, summary=summary, upto=upto)
+    db.commit()
+    log.info("chat %s compacted through %s", chat.id, upto.id)
+
+    return templates.TemplateResponse(
+        request, "chat/_thread.html", {"request": request, **_thread_context(db, chat, user)}
+    )
+
+
 @router.post("/{chat_id}/keep")
 async def keep_chat(db: Db, user: RequiredUser, chat_id: str) -> Response:
     """Stop a temporary chat being temporary.
@@ -380,6 +440,7 @@ async def _follow(chat_id: str, message_id: str) -> AsyncIterator[str]:
                 if generation.content:
                     yield sse.event("render", render_markdown(generation.text))
                 yield sse.event("metrics", _metrics_html(generation))
+                yield sse.event("status", escape_text(generation.status))
                 last_frame = time.monotonic()
 
             if generation.done:
@@ -451,16 +512,18 @@ def _metrics_html(generation) -> str:
 
 def _thread_context(db: DBSession, chat: Chat, user: User) -> dict:
     """Everything chat/_thread.html needs to render the conversation."""
-    messages = list(
+    everything = list(
         db.scalars(select(Message).where(Message.chat_id == chat.id).order_by(Message.created_at))
     )
+    compacted, messages = compaction_service.split(db, chat, everything)
     return {
         "chat": chat,
         "user": user,
         "messages": messages,
+        "compacted": compacted,
         "bodies": {
             m.id: render_markdown(m.content)
-            for m in messages
+            for m in everything
             if m.role == ROLE_ASSISTANT and m.content
         },
         "models_by_id": {m.model_id: m for m in chat_service.available_models(db, user)},
@@ -556,6 +619,14 @@ async def edit_message(
     discarded = _messages_after(db, message)
     for later in discarded:
         db.delete(later)
+
+    # A rewind to at or before the compaction boundary leaves that boundary
+    # describing turns that no longer exist. There is no foreign key to null it
+    # out on an upgraded database, so it is cleared here.
+    cutoff = compaction_service.cutoff_message(db, chat)
+    if cutoff is None or compaction_service.moment(message) <= compaction_service.moment(cutoff):
+        compaction_service.reset(chat)
+
     db.commit()
 
     assistant = chat_service.create_message(

src/lembas/api/pages.py

@@ -12,6 +12,7 @@ from lembas.db.models import Chat, Folder, KnowledgeBase, Message, User
 from lembas.security import permissions
 from lembas.services import audio as audio_service
 from lembas.services import chat as chat_service
+from lembas.services import compaction as compaction_service
 from lembas.services import settings_store
 from lembas.services import suggestions as suggestions_service
 from lembas.services.library import documents as documents_service
@@ -220,18 +221,21 @@ async def chat_detail(request: Request, db: Db, user: RequiredUser, chat_id: str
         chat.unread_notified = False
         db.commit()
 
-    messages = list(
+    everything = list(
         db.scalars(
             select(Message).where(Message.chat_id == chat.id).order_by(Message.created_at)
         )
     )
+    # Summarised turns are kept and still rendered, behind a divider -- they
+    # have only stopped being part of the request.
+    compacted, messages = compaction_service.split(db, chat, everything)
 
     # Markdown is rendered once here rather than in the template so the same
     # helper produces the page and the streamed final frame -- one code path,
     # no chance of the two disagreeing.
     bodies = {
         message.id: render_markdown(message.content)
-        for message in messages
+        for message in everything
         if message.role == "assistant" and message.content
     }
 
@@ -255,6 +259,7 @@ async def chat_detail(request: Request, db: Db, user: RequiredUser, chat_id: str
         {
             "chat": chat,
             "messages": messages,
+            "compacted": compacted,
             "bodies": bodies,
             "inherited_prompt": inherited,
             "inherited_from": inherited_from,

src/lembas/db/models/chat.py

@@ -2,9 +2,10 @@
 
 from __future__ import annotations
 
+from datetime import datetime
 from typing import TYPE_CHECKING, Any
 
-from sqlalchemy import Boolean, ForeignKey, Integer, String, Text
+from sqlalchemy import Boolean, DateTime, ForeignKey, Integer, String, Text
 from sqlalchemy.orm import Mapped, mapped_column, relationship
 
 from lembas.db.base import Base, Timestamps, UUIDPrimaryKey
@@ -108,6 +109,19 @@ class Chat(UUIDPrimaryKey, Timestamps, Base):
     unread: Mapped[bool] = mapped_column(Boolean, default=False, nullable=False)
     unread_notified: Mapped[bool] = mapped_column(Boolean, default=False, nullable=False)
 
+    # --- Compaction ----------------------------------------------------------
+    # A summary of the turns up to `compacted_through_id`, sent in their place.
+    # The messages themselves are kept and still shown; they simply stop being
+    # part of the request. See services/compaction.py.
+    compact_summary: Mapped[str] = mapped_column(Text, default="")
+    # A plain id, deliberately not a ForeignKey: db/migrations.py compiles only
+    # the column type, so a REFERENCES clause would exist on a freshly created
+    # database and not on an upgraded one, and a constraint half the fleet has
+    # is worse than none. It is validated on every read instead -- the same
+    # reasoning `model_id` above carries.
+    compacted_through_id: Mapped[str | None] = mapped_column(String(32))
+    compacted_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True))
+
     folder: Mapped[Folder | None] = relationship(back_populates="chats")
     messages: Mapped[list[Message]] = relationship(
         back_populates="chat",

src/lembas/services/chat.py

@@ -174,11 +174,31 @@ def build_messages(
     resolved, which is how the harness gets in front of the authored prompt
     without this function knowing anything about tools.
     """
+    from lembas.services import compaction as compaction_service
+    from lembas.services import prompts as prompts_service
+
     payload: list[dict[str, Any]] = []
     system = effective_system_prompt(db, chat) if system_prompt is None else system_prompt
     if system:
         payload.append({"role": ROLE_SYSTEM, "content": system})
 
+    # Compacted turns are replaced by a summary carried in two turns rather than
+    # one. A leading `assistant` breaks templates that require the first
+    # non-system message to be `user`; a lone leading `user` produces user, user
+    # whenever the kept history starts on a user turn -- which it always does,
+    # because the cutoff lands on a finished reply. The pair alternates
+    # correctly in both directions and keeps exactly one system message.
+    cutoff = compaction_service.cutoff_message(db, chat)
+    if cutoff is not None:
+        lead = prompts_service.resolve(db, "task.compact_lead").strip()
+        ack = prompts_service.resolve(db, "task.compact_ack").strip()
+        summary = chat.compact_summary.strip()
+        payload.append(
+            {"role": ROLE_USER, "content": f"{lead}\n\n{summary}" if lead else summary}
+        )
+        if ack:
+            payload.append({"role": ROLE_ASSISTANT, "content": ack})
+
     history = db.scalars(
         select(Message).where(Message.chat_id == chat.id).order_by(Message.created_at)
     ).all()
@@ -186,6 +206,10 @@ def build_messages(
     for message in history:
         if upto is not None and message.id == upto.id:
             break
+        if cutoff is not None and compaction_service.moment(
+            message
+        ) <= compaction_service.moment(cutoff):
+            continue
         # Skip turns that failed or produced nothing -- but a message carrying
         # only an attachment has no text and must still be sent.
         if message.error:
@@ -391,6 +415,42 @@ def create_message(
     return message
 
 
+async def summarise_for_compaction(
+    endpoint: Endpoint,
+    model_id: str,
+    *,
+    transcript: str,
+    previous_summary: str,
+    template: str,
+) -> str:
+    """Ask the model to summarise the earlier turns.
+
+    `template` is passed in for the same reason `generate_title`'s is: this runs
+    after the generation's session has closed, and opening another one there is
+    how you get a session that outlives its scope. An empty template means an
+    administrator cleared the fragment, and nothing is asked of anyone.
+    """
+    from lembas.services import prompts as prompts_service
+
+    if not template.strip() or not transcript.strip():
+        return ""
+
+    prompt = prompts_service.substitute(
+        template, {"transcript": transcript, "previous_summary": previous_summary}
+    )
+    raw = await complete(
+        endpoint,
+        {
+            "model": model_id,
+            "messages": [{"role": ROLE_USER, "content": prompt}],
+            "max_tokens": 1200,
+            # Low, but not zero: this is recall, not invention.
+            "temperature": 0.3,
+        },
+    )
+    return raw.strip()
+
+
 def sweep_temporary(db: DBSession, older_than: timedelta = TEMPORARY_LIFETIME) -> int:
     """Delete temporary chats nobody has touched for a day.
 

src/lembas/services/compaction.py

@@ -0,0 +1,192 @@
+"""Carrying a long conversation forward without carrying all of it.
+
+Past a certain length every chat stops working: the window fills, and the only
+options are to lose the beginning or to start again. Compaction summarises the
+earlier turns and sends the summary in their place.
+
+**The messages are kept.** They stay in the transcript, collapsed behind a
+divider, and simply stop being part of the request. A summary that turned out
+badly is then a bad turn rather than a lost conversation, which is what makes
+the button safe to press and automatic compaction safe to have at all.
+
+**Stored on the Chat, not as a synthetic Message.** A synthetic row would need a
+role: `system` breaks the one-system-message rule the moment `build_messages`
+emits it beside the harness, and `user`/`assistant` makes it a turn people can
+edit, regenerate from and copy, indistinguishable from a real one in all four
+places a bubble is rendered. Worse, "editing rewinds, it does not branch" would
+silently delete it and leave no marker that compaction had ever happened.
+"""
+
+from __future__ import annotations
+
+import logging
+from datetime import UTC, datetime
+
+from sqlalchemy import select
+from sqlalchemy.orm import Session as DBSession
+
+from lembas.db.models import ROLE_ASSISTANT, Chat, Message
+from lembas.services import metrics as metrics_service
+from lembas.services import settings_store, tokens
+
+log = logging.getLogger(__name__)
+
+# What the summariser is shown. Past this the oldest turns are dropped with a
+# marker: a transcript that does not fit the window it is protecting is no use.
+MAX_TRANSCRIPT_CHARS = 24_000
+
+# Settings key, in the GENERAL group. 0 turns automatic compaction off; the
+# button still works, because a person asking for it does not need a threshold.
+THRESHOLD_KEY = "compact_threshold"
+DEFAULT_THRESHOLD = 95
+
+
+def threshold(db: DBSession) -> int:
+    value = settings_store.get(db, THRESHOLD_KEY)
+    return int(value) if isinstance(value, (int, float)) else DEFAULT_THRESHOLD
+
+
+def moment(message: Message) -> datetime:
+    """A message's timestamp, always comparable.
+
+    SQLite does not store the offset, so a row loaded from disk comes back naive
+    while one still in the session's identity map keeps the tzinfo it was
+    created with. Comparing the two raises, and every comparison here is between
+    exactly those: a cutoff fetched by id against history loaded in bulk.
+    `files.sweep_orphans` already normalises for the same reason.
+    """
+    created = message.created_at
+    return created if created.tzinfo is not None else created.replace(tzinfo=UTC)
+
+
+def cutoff_message(db: DBSession, chat: Chat) -> Message | None:
+    """The message compaction reached, or None if it never has.
+
+    There is no foreign key to null this out on an upgraded database, so the
+    check is load-bearing rather than defensive: an id pointing at a message
+    that has been deleted means the boundary no longer describes anything, and
+    the chat has to read as uncompacted.
+    """
+    if not chat.compact_summary or not chat.compacted_through_id:
+        return None
+    message = db.get(Message, chat.compacted_through_id)
+    if message is None or message.chat_id != chat.id:
+        return None
+    return message
+
+
+def reset(chat: Chat) -> None:
+    """Forget that this chat was ever compacted."""
+    chat.compact_summary = ""
+    chat.compacted_through_id = None
+    chat.compacted_at = None
+
+
+def apply(chat: Chat, *, summary: str, upto: Message) -> None:
+    """Record a summary and move the boundary. Caller commits."""
+    chat.compact_summary = summary.strip()
+    chat.compacted_through_id = upto.id
+    chat.compacted_at = datetime.now(UTC)
+
+
+def split(
+    db: DBSession, chat: Chat, messages: list[Message]
+) -> tuple[list[Message], list[Message]]:
+    """(summarised, live) -- what is behind the divider, and what is not."""
+    cutoff = cutoff_message(db, chat)
+    if cutoff is None:
+        return [], list(messages)
+    boundary = moment(cutoff)
+    return (
+        [m for m in messages if moment(m) <= boundary],
+        [m for m in messages if moment(m) > boundary],
+    )
+
+
+def last_complete(db: DBSession, chat: Chat) -> Message | None:
+    """The newest finished assistant turn: where compaction should stop.
+
+    Landing on a reply rather than a question means the kept history starts on a
+    user turn, which is what every chat template expects.
+    """
+    return db.scalar(
+        select(Message)
+        .where(
+            Message.chat_id == chat.id,
+            Message.role == ROLE_ASSISTANT,
+            Message.complete.is_(True),
+        )
+        .order_by(Message.created_at.desc())
+    )
+
+
+def transcript(db: DBSession, chat: Chat, *, upto: Message) -> str:
+    """The turns to summarise, oldest first, as plain text.
+
+    Only the delta since the last compaction: the previous summary is supplied
+    separately, and the instruction asks for one record, so each summary
+    subsumes the one before it. Re-summarising the whole chat every time grows
+    quadratically and eventually exceeds the very window this protects.
+    """
+    previous = cutoff_message(db, chat)
+    query = select(Message).where(
+        Message.chat_id == chat.id,
+        Message.created_at <= upto.created_at,
+        Message.error == "",
+    )
+    if previous is not None:
+        query = query.where(Message.created_at > previous.created_at)
+
+    lines: list[str] = []
+    for message in db.scalars(query.order_by(Message.created_at)):
+        body = message.content.strip()
+        if not body:
+            continue
+        lines.append(f"{message.role}: {body}")
+
+    text = "\n\n".join(lines)
+    if len(text) > MAX_TRANSCRIPT_CHARS:
+        # Keep the most recent part: the older it is, the more likely the
+        # previous summary already covers it.
+        text = "[earlier turns omitted]\n\n" + text[-MAX_TRANSCRIPT_CHARS:]
+    return text
+
+
+def previous_summary_block(chat: Chat) -> str:
+    """The earlier summary, headed, or "" on a first compaction.
+
+    Empty is fine to pass straight through: `prompts.substitute` drops a line
+    that held a known variable and expanded to nothing, so the prompt does not
+    end up with a hole where a heading was.
+    """
+    if not chat.compact_summary.strip():
+        return ""
+    return "## Summary of even earlier turns\n\n" + chat.compact_summary.strip()
+
+
+def should_compact(db: DBSession, chat: Chat, *, pending: str = "") -> bool:
+    """Whether the next request should be summarised first.
+
+    Judged from the last reply's recorded usage plus an estimate of the new
+    turn. True prompt_tokens are only knowable after a response, so a
+    retrospective figure is the honest basis -- but on its own it is one turn
+    stale, and fifty thousand characters pasted into the composer would overflow
+    a window that measured 90% last time. The estimator covers only that delta.
+
+    Never fires when the model's context length is unknown. Acting on a number
+    nobody supplied is exactly what the 0-means-unknown rule exists to prevent.
+    """
+    limit = threshold(db)
+    if limit <= 0:
+        return False
+
+    last = last_complete(db, chat)
+    if last is None:
+        return False
+
+    usage = metrics_service.from_message(last.usage_json)
+    if usage.context_limit <= 0 or usage.context_tokens <= 0:
+        return False
+
+    projected = usage.context_tokens + tokens.estimate(pending)
+    return projected >= usage.context_limit * limit / 100

src/lembas/services/generation.py

@@ -22,9 +22,12 @@ import time
 from dataclasses import dataclass, field
 from datetime import UTC, datetime, timedelta
 
+from sqlalchemy import select
+
 from lembas.db.models import ROLE_ASSISTANT, ROLE_USER, Chat, Message, User
 from lembas.db.session import session_scope
 from lembas.services import chat as chat_service
+from lembas.services import compaction as compaction_service
 from lembas.services import metrics as metrics_service
 from lembas.services import prompts as prompts_service
 from lembas.services import tokens
@@ -95,6 +98,10 @@ class Generation:
     # woken individually: with a 100ms cadence a short poll is simpler than
     # future bookkeeping, and cannot drop a wakeup.
     version: int = 0
+    # What the reply is doing when it is not producing tokens. Shown in the
+    # streaming bubble, because a silent multi-second pause before the first
+    # token is what a hang looks like.
+    status: str = ""
     # Number of browsers currently watching. Decides whether a finished reply
     # counts as unread.
     followers: int = 0
@@ -214,6 +221,13 @@ async def _run(generation: Generation) -> None:
     title_prompt = ""
 
     try:
+        # Before the request is assembled, so build_request is called once and
+        # what goes out is the compacted conversation -- there is no second
+        # assembly path. Here rather than in post_message because that route's
+        # whole contract is to return immediately, and a three-second
+        # summarisation in front of it would break exactly that.
+        await _maybe_compact(generation)
+
         with session_scope() as db:
             chat = db.get(Chat, generation.chat_id)
             message = db.get(Message, generation.message_id)
@@ -387,6 +401,77 @@ async def _run(generation: Generation) -> None:
         generation.touch()
 
 
+async def _maybe_compact(generation: Generation) -> None:
+    """Summarise the earlier turns if the window is about to be full.
+
+    Never raises. A failed compaction logs and sends the uncompacted request,
+    which either works or fails upstream with a message that says what actually
+    happened -- refusing to answer because the summariser was unavailable would
+    be a worse trade.
+
+    The awaited call is deliberately outside any session, the same shape titling
+    uses: read everything needed, close, ask, reopen to write.
+    """
+    try:
+        with session_scope() as db:
+            chat = db.get(Chat, generation.chat_id)
+            message = db.get(Message, generation.message_id)
+            if chat is None or message is None:
+                return
+
+            pending = _pending_text(db, message)
+            if not compaction_service.should_compact(db, chat, pending=pending):
+                return
+
+            template = prompts_service.resolve(db, "task.compact")
+            upto = compaction_service.last_complete(db, chat)
+            if not template.strip() or upto is None:
+                return
+
+            endpoint, model_id = chat_service.resolve_endpoint(db, chat)
+            transcript = compaction_service.transcript(db, chat, upto=upto)
+            previous = compaction_service.previous_summary_block(chat)
+            upto_id = upto.id
+
+        generation.status = "Summarising earlier messages…"
+        generation.touch()
+
+        summary = await chat_service.summarise_for_compaction(
+            endpoint,
+            model_id,
+            transcript=transcript,
+            previous_summary=previous,
+            template=template,
+        )
+        if not summary:
+            return
+
+        with session_scope() as db:
+            chat = db.get(Chat, generation.chat_id)
+            upto = db.get(Message, upto_id)
+            if chat is None or upto is None:
+                return
+            compaction_service.apply(chat, summary=summary, upto=upto)
+            db.commit()
+            log.info("chat %s compacted automatically through %s", chat.id, upto_id)
+    except Exception:  # noqa: BLE001 - the reply matters more than the tidy-up
+        log.exception("automatic compaction failed for chat %s", generation.chat_id)
+    finally:
+        generation.status = ""
+        generation.touch()
+
+
+def _pending_text(db, message: Message) -> str:
+    """The user turn this reply is answering, for the size estimate."""
+    previous = db.scalars(
+        select(Message)
+        .where(Message.chat_id == message.chat_id, Message.created_at < message.created_at)
+        .order_by(Message.created_at.desc())
+        .limit(1)
+    ).first()
+    return previous.content if previous is not None else ""
+
+
 def _question_from(payload: dict) -> str:
     """The last thing the user said, for auto-titling."""
     for entry in reversed(payload.get("messages", [])):

src/lembas/services/prompts.py

@@ -156,6 +156,16 @@ VARIABLES: tuple[Variable, ...] = (
     ),
     Variable("question", "Question", "The first message. Chat title task only."),
     Variable("answer", "Answer", "The first reply. Chat title task only."),
+    Variable(
+        "transcript",
+        "Transcript",
+        "The turns being summarised, oldest first. Compaction task only.",
+    ),
+    Variable(
+        "previous_summary",
+        "Earlier summary",
+        "The summary from a previous compaction, if there was one. Compaction task only.",
+    ),
 )
 
 VARIABLE_NAMES = frozenset(variable.name for variable in VARIABLES)
@@ -764,6 +774,77 @@ BUILTIN: tuple[Fragment, ...] = (
             "Assistant: {{answer}}"
         ),
     ),
+    Fragment(
+        key="task.compact",
+        label="Compaction summary",
+        group=GROUP_TASKS,
+        order=410,
+        variables=("transcript", "previous_summary"),
+        hint="A separate one-message request, not part of any chat. Clear it to "
+        "turn compaction off entirely: the button says so and nothing is "
+        "summarised automatically.",
+        default=(
+            "Summarise the conversation below so it can be carried forward after the "
+            "earlier turns are dropped from your context. This is a working record, "
+            "not a report for a reader.\n"
+            "\n"
+            "Keep, under these headings and in this order:\n"
+            "\n"
+            "## What we are doing\n"
+            "The goal, and where we have got to.\n"
+            "\n"
+            "## Decisions\n"
+            "Anything settled, and why. A decision without its reason gets argued "
+            "again.\n"
+            "\n"
+            "## Facts established\n"
+            "Names, numbers, versions, file paths, URLs and identifiers, copied "
+            "exactly. Do not round them, paraphrase them or reconstruct one from "
+            "memory — if it is not in the transcript, leave it out.\n"
+            "\n"
+            "## Open threads\n"
+            "What is unfinished, and what was about to happen next.\n"
+            "\n"
+            "Leave out pleasantries, retracted ideas and anything already superseded. "
+            "Do not answer the conversation: you are recording it. Write in the "
+            "language of the conversation, and stay under 500 words.\n"
+            "\n"
+            "{{previous_summary}}\n"
+            "\n"
+            "## Transcript\n"
+            "\n"
+            "{{transcript}}"
+        ),
+    ),
+    Fragment(
+        key="task.compact_lead",
+        label="How a summary is introduced",
+        group=GROUP_TASKS,
+        order=420,
+        hint="Sits in front of the summary, in the turn that replaces the "
+        "messages no longer being sent. Without it a model reads the summary as "
+        "something the person has just typed.",
+        default=(
+            "Here is a summary of the earlier part of this conversation. Those "
+            "messages are no longer in your context. Treat this summary as an "
+            "accurate record of them and rely on it rather than on what you can no "
+            "longer see; if it does not cover something you need, say so instead of "
+            "filling the gap."
+        ),
+    ),
+    Fragment(
+        key="task.compact_ack",
+        label="The model's acknowledgement",
+        group=GROUP_TASKS,
+        order=430,
+        hint="One assistant turn after the summary, so the conversation still "
+        "alternates user, assistant, user. Several chat templates reject a "
+        "history that does not.",
+        default=(
+            "Understood. I have the summary of the earlier turns and will carry on "
+            "from there."
+        ),
+    ),
 )
 
 register_source(_builtin_source)

src/lembas/services/settings_store.py

@@ -35,6 +35,12 @@ def _general_defaults() -> dict[str, Any]:
         # Applied to every chat that has no model or chat prompt of its
         # own. See services.chat.effective_system_prompt.
         "system_prompt": "",
+        # Percentage of a model's context length at which the earlier turns are
+        # summarised automatically. 0 turns it off; the Compact button still
+        # works, because a person asking for it does not need a threshold.
+        # Never fires for a model whose context_length is 0, since that is
+        # "unknown" rather than "small". See services/compaction.py.
+        "compact_threshold": 95,
     }
 
 

src/lembas/web/static/css/chat.css

@@ -167,6 +167,45 @@
 .reasoning__summary::-webkit-details-marker { display: none; }
 .reasoning__summary:hover { color: var(--ink); background: var(--surface-hover); }
 
+/* --- Compacted turns -------------------------------------------------------
+   Summarised messages, kept and readable but out of the way. Collapsed by
+   default: the point of compacting was that they stopped mattering.
+*/
+.compacted {
+  border: 1px dashed var(--border-strong);
+  border-radius: var(--radius);
+  background: color-mix(in srgb, var(--surface) 50%, transparent);
+}
+.compacted__summary {
+  display: flex;
+  align-items: center;
+  gap: var(--sp-2);
+  padding: var(--sp-2) var(--sp-3);
+  cursor: pointer;
+  color: var(--ink-muted);
+  font-size: var(--text-sm);
+  list-style: none;
+  user-select: none;
+}
+.compacted__summary::-webkit-details-marker { display: none; }
+.compacted__summary:hover { color: var(--ink); }
+.compacted[open] .reasoning__chevron { transform: rotate(180deg); }
+
+.compacted__note {
+  margin: 0;
+  padding: 0 var(--sp-3);
+  font-size: var(--text-xs);
+  color: var(--ink-faint);
+  line-height: var(--leading-normal);
+}
+.compacted__body {
+  display: flex;
+  flex-direction: column;
+  gap: var(--sp-6);
+  padding: var(--sp-4) var(--sp-3);
+  opacity: 0.75;
+}
+
 /* --- Suggestions -----------------------------------------------------------
    Starting points on the empty screen. Cards rather than a list, because they
    are things to press.
@@ -333,6 +372,9 @@
 }
 
 /* --- Stop, notes and editing ---------------------------------------------- */
+.msg__status { font-size: var(--text-xs); color: var(--ink-faint); font-style: italic; }
+.msg__status:empty { display: none; }
+
 .msg__waiting {
   display: flex;
   align-items: center;

src/lembas/web/templates/admin/general.html

@@ -41,6 +41,29 @@
     </div>
   </section>
 
+  <section class="card">
+    <h2 class="card__title">Compaction</h2>
+    <p class="card__lede">
+      A long conversation eventually fills the model's context. When it gets
+      close, the earlier turns are summarised and the summary is sent in their
+      place. The messages themselves are kept and stay readable in the
+      transcript — they simply stop being sent.
+    </p>
+    <div class="field">
+      <label class="field__label" for="compact-threshold">Compact at</label>
+      <input class="input" id="compact-threshold" name="compact_threshold" type="number"
+             min="0" max="99" value="{{ values.compact_threshold }}">
+      <p class="field__hint">
+        Percent of the model's context length. <code>0</code> turns automatic
+        compaction off; the button in each chat still works. Nothing happens for
+        a model whose context length is unset under
+        <a href="/admin/models">Models</a> — that is "unknown", not "small", and
+        this will not act on a number nobody supplied. The wording of the
+        summary is under <a href="/admin/prompts">Prompts</a>.
+      </p>
+    </div>
+  </section>
+
   <section class="card">
     <h2 class="card__title">
       Registration

src/lembas/web/templates/chat/_message.html

@@ -119,6 +119,10 @@
          reply is being written, which is where the hand already is. #}
       <div class="msg__waiting">
         <span class="dots"><i></i><i></i><i></i></span>
+        {# What the reply is doing when it is not producing tokens. A silent
+           multi-second pause before the first token is what a hang looks
+           like. #}
+        <span class="msg__status" sse-swap="status" hx-swap="innerHTML"></span>
       </div>
       {# Counts as the reply is written. Everything is an estimate until the
          usage chunk lands at the very end, and the chips say so. #}

src/lembas/web/templates/chat/_thread.html

@@ -1,12 +1,38 @@
+{% from "_macros.html" import icon %}
 {#
-  The whole thread. Returned after a rewind, which changes an arbitrary number
-  of messages at once -- replacing the lot is simpler and less error-prone than
-  working out which individual bubbles to remove. Also included by
-  chat/index.html, so the conversation is described in exactly one place.
+  The whole thread. Returned after a rewind or a compaction, either of which
+  changes an arbitrary number of messages at once -- replacing the lot is
+  simpler and less error-prone than working out which individual bubbles to
+  remove. Also included by chat/index.html, so the conversation is described in
+  exactly one place.
 
   Deliberately has no root element: it is swapped with innerHTML into #thread,
   and an outerHTML swap would take the container with it.
 #}
+{% if compacted %}
+{# Summarised turns are kept and still readable -- they have only stopped being
+   sent. A compaction that summarised badly is then a bad turn rather than a
+   lost conversation. #}
+<details class="compacted">
+  <summary class="compacted__summary">
+    {{ icon("archive", "icon--sm") }}
+    <span>{{ compacted | length }} earlier messages, summarised</span>
+    {{ icon("chevron-down", "icon--sm reasoning__chevron") }}
+  </summary>
+  <p class="compacted__note">
+    These are no longer sent to the model; a summary of them goes instead. They
+    are kept here so nothing is lost.
+  </p>
+  <div class="compacted__body">
+    {% for message in compacted %}
+      {% with body_html = bodies.get(message.id, "") %}
+        {% include "chat/_message.html" %}
+      {% endwith %}
+    {% endfor %}
+  </div>
+</details>
+{% endif %}
+
 {% for message in messages %}
   {% with body_html = bodies.get(message.id, "") %}
     {% include "chat/_message.html" %}

src/lembas/web/templates/chat/index.html

@@ -66,6 +66,18 @@
         </button>
         {% endif %}
 
+        {% if chat and messages %}
+        <button class="btn btn--icon" type="button" aria-label="Compact this chat"
+                title="Summarise the earlier messages so they stop taking up context"
+                hx-post="/api/chats/{{ chat.id }}/compact"
+                hx-target="#thread" hx-swap="innerHTML"
+                hx-confirm="Summarise everything before the last reply? The messages stay in the transcript; they just stop being sent to the model."
+                data-confirm-title="Compact this chat"
+                data-confirm-label="Compact">
+          {{ icon("archive") }}
+        </button>
+        {% endif %}
+
         {% if chat and user.is_admin %}
         <button class="btn btn--icon" type="button" aria-label="Inspect this chat"
                 title="Inspect this chat" aria-expanded="false" data-toggle="#inspector">

tests/test_compaction.py

@@ -0,0 +1,372 @@
+"""Compaction: what is summarised, what is sent, and when it happens by itself."""
+
+from __future__ import annotations
+
+import json
+from datetime import UTC, datetime, timedelta
+
+import httpx
+import pytest
+from fastapi.testclient import TestClient
+from sqlalchemy import select
+
+from lembas.db.models import Chat, Connection, Message, Model
+from lembas.services import chat as chat_service
+from lembas.services import compaction as compaction_service
+from lembas.services import prompts, settings_store
+from lembas.services.crypto import encrypt
+
+
+@pytest.fixture
+def chat(db, registered, make_chat) -> Chat:
+    connection = Connection(
+        name="Test", base_url="http://x.test", api_key_encrypted=encrypt("")
+    )
+    db.add(connection)
+    db.commit()
+    db.add(
+        Model(connection_id=connection.id, model_id="test-model", context_length=1000)
+    )
+    db.commit()
+    return db.get(Chat, make_chat())
+
+
+def _exchange(db, chat: Chat, *, question: str, answer: str, minutes: int = 0) -> Message:
+    """One user turn and its reply, backdated so ordering is deterministic."""
+    when = datetime.now(UTC) - timedelta(minutes=minutes)
+    db.add(Message(chat_id=chat.id, role="user", content=question, created_at=when))
+    reply = Message(
+        chat_id=chat.id,
+        role="assistant",
+        content=answer,
+        created_at=when + timedelta(seconds=1),
+    )
+    db.add(reply)
+    db.commit()
+    return reply
+
+
+def _usage(reply: Message, *, context_tokens: int, limit: int = 1000) -> None:
+    reply.usage_json = {
+        "prompt_tokens": context_tokens,
+        "completion_tokens": 0,
+        "total_tokens": context_tokens,
+        "context_tokens": context_tokens,
+        "context_limit": limit,
+        "estimated": False,
+        "elapsed_ms": 10,
+        "rounds": 1,
+    }
+
+
+# --- The boundary -------------------------------------------------------------
+def test_an_uncompacted_chat_splits_into_nothing_and_everything(db, chat):
+    _exchange(db, chat, question="one", answer="two")
+    messages = list(db.scalars(select(Message).order_by(Message.created_at)))
+    assert compaction_service.split(db, chat, messages) == ([], messages)
+
+
+def test_a_dangling_cutoff_reads_as_uncompacted(db, chat):
+    """There is no foreign key to null it out on an upgraded database, so the
+    guard is load-bearing rather than defensive."""
+    reply = _exchange(db, chat, question="one", answer="two", minutes=10)
+    compaction_service.apply(chat, summary="a summary", upto=reply)
+    db.commit()
+
+    db.delete(reply)
+    db.commit()
+
+    assert compaction_service.cutoff_message(db, chat) is None
+    assert compaction_service.split(db, chat, [])[0] == []
+
+
+def test_the_cutoff_lands_on_a_finished_reply(db, chat):
+    """So the kept history starts on a user turn, which is what every chat
+    template expects."""
+    _exchange(db, chat, question="one", answer="two", minutes=10)
+    db.add(Message(chat_id=chat.id, role="user", content="three"))
+    db.commit()
+
+    assert compaction_service.last_complete(db, chat).content == "two"
+
+
+# --- What gets sent -----------------------------------------------------------
+def test_the_summary_replaces_the_compacted_turns(db, chat):
+    old = _exchange(db, chat, question="what is lembas?", answer="Waybread.", minutes=10)
+    _exchange(db, chat, question="how much?", answer="A bite.", minutes=1)
+    compaction_service.apply(chat, summary="They asked about lembas.", upto=old)
+    db.commit()
+
+    messages = chat_service.build_messages(db, chat)
+    contents = [m["content"] for m in messages]
+
+    assert "what is lembas?" not in contents
+    assert "how much?" in contents
+    assert any("They asked about lembas." in c for c in contents)
+
+
+def test_the_summary_is_carried_by_a_user_and_an_assistant_turn(db, chat):
+    """A leading assistant turn breaks templates requiring the first non-system
+    message to be user; a lone leading user turn produces user, user whenever the
+    kept history starts on a user turn -- which it always does."""
+    old = _exchange(db, chat, question="one", answer="two", minutes=10)
+    _exchange(db, chat, question="three", answer="four", minutes=1)
+    compaction_service.apply(chat, summary="Summary.", upto=old)
+    db.commit()
+
+    roles = [m["role"] for m in chat_service.build_messages(db, chat)]
+    assert roles[:2] == ["user", "assistant"]
+    # And it still alternates from there.
+    assert roles == ["user", "assistant", "user", "assistant"]
+
+
+def test_there_is_still_exactly_one_system_message(db, chat):
+    settings_store.update(db, {"system_prompt": "Speak as Gandalf."})
+    old = _exchange(db, chat, question="one", answer="two", minutes=10)
+    compaction_service.apply(chat, summary="Summary.", upto=old)
+    db.commit()
+
+    roles = [m["role"] for m in chat_service.build_messages(db, chat, system_prompt="S")]
+    assert roles.count("system") == 1
+    assert roles[0] == "system"
+
+
+def test_clearing_the_lead_fragment_still_sends_the_summary(db, chat):
+    prompts.save(db, {"task.compact_lead": "", "task.compact_ack": ""})
+    old = _exchange(db, chat, question="one", answer="two", minutes=10)
+    compaction_service.apply(chat, summary="Summary.", upto=old)
+    db.commit()
+
+    messages = chat_service.build_messages(db, chat)
+    assert messages[0] == {"role": "user", "content": "Summary."}
+
+
+# --- The transcript -----------------------------------------------------------
+def test_only_the_delta_is_summarised_the_second_time(db, chat):
+    """Re-summarising the whole chat grows quadratically and eventually exceeds
+    the very window it is protecting."""
+    first = _exchange(db, chat, question="the old part", answer="ok", minutes=20)
+    compaction_service.apply(chat, summary="Earlier summary.", upto=first)
+    db.commit()
+
+    second = _exchange(db, chat, question="the new part", answer="ok", minutes=5)
+    transcript = compaction_service.transcript(db, chat, upto=second)
+
+    assert "the new part" in transcript
+    assert "the old part" not in transcript
+    assert "Earlier summary." in compaction_service.previous_summary_block(chat)
+
+
+def test_a_first_compaction_has_no_previous_summary(db, chat):
+    assert compaction_service.previous_summary_block(chat) == ""
+
+
+def test_a_huge_transcript_is_trimmed_from_the_front(db, chat):
+    reply = _exchange(db, chat, question="x" * 40_000, answer="ok", minutes=5)
+    transcript = compaction_service.transcript(db, chat, upto=reply)
+
+    assert len(transcript) < 40_000
+    assert transcript.startswith("[earlier turns omitted]")
+
+
+# --- The button ---------------------------------------------------------------
+def _summariser(text: str = "## What we are doing\n\nAsking about lembas."):
+    def handler(request: httpx.Request) -> httpx.Response:
+        return httpx.Response(200, json={"choices": [{"message": {"content": text}}]})
+
+    return handler
+
+
+async def test_compacting_summarises_and_hides_the_earlier_turns(
+    client: TestClient, db, chat, mock_http
+):
+    _exchange(db, chat, question="what is lembas?", answer="Waybread.", minutes=10)
+    mock_http(_summariser())
+
+    response = client.post(f"/api/chats/{chat.id}/compact")
+    assert response.status_code == 200
+
+    db.refresh(chat)
+    assert "Asking about lembas." in chat.compact_summary
+    assert chat.compacted_through_id
+    # The messages are still there, behind the divider.
+    assert "earlier messages, summarised" in response.text
+    assert db.scalar(select(Message).where(Message.content == "what is lembas?")) is not None
+
+
+async def test_compacting_while_a_reply_is_being_written_is_refused(
+    client: TestClient, db, chat, mock_http
+):
+    _exchange(db, chat, question="one", answer="two", minutes=10)
+    db.add(Message(chat_id=chat.id, role="assistant", content="", complete=False))
+    db.commit()
+
+    response = client.post(f"/api/chats/{chat.id}/compact")
+    assert response.status_code == 409
+
+
+async def test_compaction_can_be_turned_off_by_clearing_its_prompt(
+    client: TestClient, db, chat, mock_http
+):
+    _exchange(db, chat, question="one", answer="two", minutes=10)
+    prompts.save(db, {"task.compact": ""})
+
+    response = client.post(f"/api/chats/{chat.id}/compact")
+    assert response.status_code == 409
+    assert "turned off" in response.json()["detail"]
+
+
+async def test_compacting_an_empty_chat_is_refused(client: TestClient, db, chat, mock_http):
+    assert client.post(f"/api/chats/{chat.id}/compact").status_code == 409
+
+
+# --- Rewinding across the boundary --------------------------------------------
+def test_editing_at_or_before_the_cutoff_clears_the_compaction(
+    client: TestClient, db, chat, mock_http
+):
+    """A rewind deletes everything after the edited message, so a boundary at or
+    behind it no longer describes anything that exists."""
+    first_reply = _exchange(db, chat, question="one", answer="two", minutes=20)
+    _exchange(db, chat, question="three", answer="four", minutes=10)
+    compaction_service.apply(chat, summary="Summary.", upto=first_reply)
+    db.commit()
+
+    first_user = db.scalar(select(Message).where(Message.content == "one"))
+    mock_http(_summariser())
+    client.post(
+        f"/api/chats/{chat.id}/messages/{first_user.id}/edit", data={"content": "one again"}
+    )
+
+    db.refresh(chat)
+    assert chat.compact_summary == ""
+    assert chat.compacted_through_id is None
+
+
+# --- Automatic ----------------------------------------------------------------
+def test_it_fires_when_the_window_is_nearly_full(db, chat):
+    reply = _exchange(db, chat, question="one", answer="two", minutes=5)
+    _usage(reply, context_tokens=960)
+    db.commit()
+
+    assert compaction_service.should_compact(db, chat) is True
+
+
+def test_it_does_not_fire_with_room_to_spare(db, chat):
+    reply = _exchange(db, chat, question="one", answer="two", minutes=5)
+    _usage(reply, context_tokens=400)
+    db.commit()
+
+    assert compaction_service.should_compact(db, chat) is False
+
+
+def test_a_large_pending_turn_is_counted(db, chat):
+    """The recorded figure is one turn stale. Fifty thousand characters pasted
+    into the composer overflow a window that measured 90% last time."""
+    reply = _exchange(db, chat, question="one", answer="two", minutes=5)
+    _usage(reply, context_tokens=900)
+    db.commit()
+
+    assert compaction_service.should_compact(db, chat) is False
+    assert compaction_service.should_compact(db, chat, pending="x" * 400) is True
+
+
+def test_it_never_fires_without_a_context_length(db, chat):
+    """Acting on a number nobody supplied is exactly what 0-means-unknown is
+    there to prevent."""
+    reply = _exchange(db, chat, question="one", answer="two", minutes=5)
+    _usage(reply, context_tokens=99_000, limit=0)
+    db.commit()
+
+    assert compaction_service.should_compact(db, chat) is False
+
+
+def test_a_threshold_of_zero_turns_it_off(db, chat):
+    settings_store.update(db, {"compact_threshold": 0})
+    reply = _exchange(db, chat, question="one", answer="two", minutes=5)
+    _usage(reply, context_tokens=999)
+    db.commit()
+
+    assert compaction_service.should_compact(db, chat) is False
+
+
+def test_it_does_not_fire_on_the_first_turn(db, chat):
+    assert compaction_service.should_compact(db, chat) is False
+
+
+def test_it_acts_on_estimated_counts_too(db, chat):
+    """A premature compaction costs one turn of answer quality, not data -- the
+    messages are still there. That is what makes acting on an estimate safe."""
+    reply = _exchange(db, chat, question="one", answer="two", minutes=5)
+    _usage(reply, context_tokens=960)
+    reply.usage_json = {**reply.usage_json, "estimated": True}
+    db.commit()
+
+    assert compaction_service.should_compact(db, chat) is True
+
+
+async def test_a_generation_compacts_before_it_asks(db, chat, mock_http):
+    """At the top of _run, so build_request is called once and what goes out is
+    the compacted conversation."""
+    from lembas.services import generation as generation_service
+
+    reply = _exchange(db, chat, question="what is lembas?", answer="Waybread.", minutes=10)
+    _usage(reply, context_tokens=980)
+    db.commit()
+
+    pending = Message(chat_id=chat.id, role="user", content="more?")
+    db.add(pending)
+    db.commit()
+    placeholder = Message(chat_id=chat.id, role="assistant", content="", complete=False)
+    db.add(placeholder)
+    db.commit()
+
+    sent: list[dict] = []
+
+    def handler(request: httpx.Request) -> httpx.Response:
+        body = json.loads(request.content)
+        sent.append(body)
+        if body.get("stream"):
+            return httpx.Response(200, text="data: [DONE]\n\n")
+        return httpx.Response(200, json={"choices": [{"message": {"content": "A summary."}}]})
+
+    mock_http(handler)
+    generation = generation_service.Generation(chat_id=chat.id, message_id=placeholder.id)
+    generation_service._RUNNING[placeholder.id] = generation
+    await generation_service._run(generation)
+    generation_service._RUNNING.clear()
+
+    db.expire_all()
+    assert db.get(Chat, chat.id).compact_summary == "A summary."
+    # The streamed request went out after compaction, carrying the summary.
+    streamed = next(b for b in sent if b.get("stream"))
+    assert any("A summary." in str(m.get("content")) for m in streamed["messages"])
+    assert not any(m.get("content") == "what is lembas?" for m in streamed["messages"])
+
+
+async def test_a_failed_compaction_still_sends_the_reply(db, chat, mock_http):
+    """Refusing to answer because the summariser was unavailable is a worse
+    trade than sending the uncompacted request."""
+    from lembas.services import generation as generation_service
+
+    reply = _exchange(db, chat, question="one", answer="two", minutes=10)
+    _usage(reply, context_tokens=980)
+    db.commit()
+    placeholder = Message(chat_id=chat.id, role="assistant", content="", complete=False)
+    db.add(placeholder)
+    db.commit()
+
+    def handler(request: httpx.Request) -> httpx.Response:
+        if json.loads(request.content).get("stream"):
+            return httpx.Response(
+                200, text='data: {"choices":[{"delta":{"content":"hi"}}]}\n\ndata: [DONE]\n\n'
+            )
+        return httpx.Response(500, json={"error": {"message": "no"}})
+
+    mock_http(handler)
+    generation = generation_service.Generation(chat_id=chat.id, message_id=placeholder.id)
+    generation_service._RUNNING[placeholder.id] = generation
+    await generation_service._run(generation)
+    generation_service._RUNNING.clear()
+
+    assert generation.text == "hi"
+    assert not generation.error