powerMonitor's AC/battery state is mirrored to the renderers
(store/power.ts) and visiblePoll quadruples its cadence on battery. Only
the safety-net refreshes slow down — event-driven refreshes and live
streaming are untouched.
It loads arbitrary user-linked pages offscreen; unthrottled, a heavy page
burns full CPU for the window's whole lifetime. Title resolution rides
load events and main-process timers, which throttling doesn't touch.
The process-wide disable-background-timer-throttling /
disable-backgrounding-occluded-windows switches plus a static
backgroundThrottling: false on every chat window pinned each renderer's
document.visibilityState to 'visible' for the life of the window. Every
visibility-gated backstop poll and clock tick in the renderer became an
always-on timer: an idle, minimized Hermes burned ~20% CPU around the
clock, on battery too.
Throttling is now a runtime dial. A small controller (stream-throttle.ts)
rides the merged hermes:active-work reports the quit guard already
receives: while any turn is in flight every chat window gets
setBackgroundThrottling(false) — a live answer keeps painting while
blurred, occluded, or minimized, exactly as before — and once all turns
settle (plus a 5s trailing window so the final flush lands at full
cadence) Chromium's default throttling returns and hidden windows go
quiet.
disable-renderer-backgrounding stays: process priority only, no timer
semantics, and it keeps hidden streaming fast.
Org-shared skills were unusable past the first propose. Three defects, one
root cause plus two that it masked.
ROOT CAUSE — org reads went to the personal endpoint.
`SyncClient.get_refs()` / `get_object()` only ever called `/v1/sync/refs`
and `/v1/sync/objects/:hash`. Those routes are hard-scoped server-side to
the token's own owner, so asking them for `refs/org/<id>/` returns the
caller's PERSONAL refs rather than an error, and org objects 404. Both org
call sites read org state through them:
- `pull_org_skills` resolved head=None for a populated org and reported
`{"ok": true, "head": null, "updated": []}` — org skills silently never
arrived, which reads as "my org has no skills" rather than as a failure.
- `propose_skill` resolved base_head=None, so the FIRST propose to an org
succeeded by accident (`from: null` happened to be correct) and EVERY
later one CAS'd against a head it had never seen -> 409 -> a raw
`SyncConflict` traceback. Worse, it built its root from an empty skill
map, so a landed CAS would have REPLACED the org set rather than splicing
into it — the 409 was accidentally preventing data loss.
Fix: `org_scope=True` on `get_refs`/`get_object`, threaded through
`get_commit_json`, `get_tree_json`, `_root_tree_of_commit`,
`_skill_trees_of_root`, and `materialize_tree` — walking an org commit needs
the org route on every hop, not just the first. Both org call sites now go
through one `_read_org_head()` helper.
ALSO FIXED
- `propose_skill` retries on conflict. When the org HEAD moves between the
read and the CAS (another member proposing, an admin merging), it
re-splices this one skill onto the NEW head and retries, bounded at 5
attempts. Re-splicing rather than replaying the old root is what stops a
concurrent proposal being dropped.
- An empty `actual` in a 409 means "the ref does not exist", not "here is a
commit". `SyncConflict` normalizes "" to None in its constructor, and the
personal push path redoes the CAS as a create instead of fetching "" as an
object — which surfaced as the baffling `object not found` (doubled
space). This is what a client hits after switching sync planes, since
`.sync_state` is not environment-scoped and carries a foreign head.
THE MOCK WAS THE REASON THIS SHIPPED
The test mock served org refs and org objects off the personal routes, so
21 org tests passed against a client that could not work against the real
plane. The mock now mirrors production: `/v1/sync/org/refs` and
`/v1/sync/org/objects/:hash` exist, org objects live in a separate scope,
and the personal routes refuse org content. Two existing tests had to be
corrected to assert against the org scope — they had been passing on the
mock's over-permissiveness.
Tests: 5 new (org head invisible on the personal route; second propose
splices and preserves the first; pull resolves a real org head; empty
`actual` -> None; push recovers from a stale cross-plane head). Verified
they FAIL without the fix: reverting just `_read_org_head` to the personal
route fails the second-propose test and the pre-existing splice test.
1278 passed / 0 failed across 54 suites via scripts/run_tests.sh.
Verified against PRODUCTION with a real org token, not just the mock:
- `pull_org_skills` -> head `sha256:1adf9333…`, materialized
`software-development/gateway-gateway-connector` into the `_org` mirror
(was head=None, updated=[]).
- A second `hermes sync propose` succeeded where it previously raised, and
the org set afterwards contains BOTH skills with the new commit
descending from the first.
The token in the input line is the whole receipt. Drop the notices that
duplicated it somewhere the user was not looking: the drag-drop and
clipboard sys() lines, and the attachedImageNotice / "detected file: X"
activity rows above the status bar.
attachedImageNotice and imageTokenMeta have no callers left.
Every attach path now drops an `[[ Image N ]]` token where you are typing:
drag-drop, clipboard (bracketed and hotkey), /image, /paste. The composer
owns clipboard attach directly instead of calling back out to useMainApp.
Deleting the token is how you unattach — there is no second control.
updateInput is the one choke point every keystroke passes through, so
syncTokens reconciles there and detaches anything erased. That also fixes
a stale image riding along on the next unrelated turn.
Tokens and the input line get refs alongside state: paste-then-immediately
-Enter submits before React has re-rendered, and the submit path has to see
the token that was just added.
A collapsed paste and an attached image are the same idea: a `[[ … ]]`
marker sitting in the input line that stands in for a payload resolved at
submit. Model both as ComposerToken and give them one expander.
Image tokens resolve to nothing — the gateway already holds the file in
attached_images — so expandTokens eats an adjacent space to avoid leaving
a gap mid-sentence. nextImageIndex never reuses an index after a delete,
or two files would collide on one label.
Select-all + Cut emptied the text and left the composer blank — no draft,
no prompt. Delete had the same hole.
The placeholder is painted on `:empty`, and a cleared editor keeps a
scaffolding <br> so the contenteditable can't collapse to a sliver. Those
two facts collide: the moment the break lands the editor has a child,
`:empty` goes false, and the prompt never comes back.
CSS can't infer emptiness on its own either. A text node is invisible to
selectors, so `one<br>` and a lone `<br>` are the same shape — a structural
rule like `:has(> br:only-child)` paints the placeholder straight over the
user's text. The code that empties the editor is what knows, so it marks
the root and the condition reads `:is(:empty, [data-empty])`.
Both writers that reshape that root maintain the marker through one helper:
the normalizer, and renderComposerContents for a restored draft or an undo.
The message-edit composer shares the slot and the rule, so it takes the
same shared class instead of drifting on its own copy.
#74815 fixed the draft this stashed; the placeholder is a separate seam.
The tab strip decided the close gesture from the `uncloseable` flag, which the
workspace sets to keep its pane in the tree — so the one tab whose close now
does something couldn't be ⌘-clicked or middle-clicked, and its right-click
menu had no Close.
Read the gesture off the pane's registered closer instead, with the workspace
registering closeWorkspaceTab. An atom rather than a lookup, since that closer
comes from a wiring effect that lands after the strip's first paint.
The workspace pane can't leave the tree, so "close the main tab" only ever had
one answer wired: shift the next stacked session in. With main as the only tab
there was nothing to shift and ⌘W dead-ended on the tab the user was looking
at.
closeWorkspaceTab is now the one answer for every entry point — stacked
session still wins, and with nothing stacked main drops to a fresh New session
draft. A blank draft and a full-page view stay no-ops: a blank draft already
IS the post-close state.
Chromium on Windows and Linux answers a middle press inside a scroller by
starting the autoscroll pan, and the mouseup that ends the pan never becomes
an auxclick. Every surface carrying the gesture — tab strips, the session
list, the terminal rail — is a scroller, so middle-click only ever worked on
macOS, where autoscroll doesn't exist.
Arm on pointerdown, spend on the pointerup over the same element (press one
tab, release on another and nothing happens), and cancel the middle mousedown
on every press so the pan widget can't appear on a surface that owns the
button. One helper, four call sites.
⌥-drag is the app's force-selection gesture over mouse-mode TUIs, but
xterm's default alt-click-moves-cursor claims the same click and emits one
cursor left/right escape per column of travel. Shells that don't consume
them echo the raw `^[[D` burst into the buffer. One gesture, one meaning.
Both of xterm's link paths activate through `window.open()`, which the
window's setWindowOpenHandler denies, so ⌘-clicking a URL did nothing but
log "Opening link blocked as opener could not be cleared" — and the OSC 8
path fronted that dead end with a raw confirm() dialog. Route both through
the desktop bridge, the path every other external link in the app takes.
⌘-click on macOS, Ctrl-click elsewhere, matching VS Code's integrated
terminal, Terminal.app, and iTerm2. A bare click stays with the selection so
a misclick on a URL can't launch a browser.
The two exact-list assertions in test_tui_gateway_server froze the toolset
list and broke the moment _RECENTLY_SHIPPED_TOOLSETS back-filled bfl onto a
saved platform list — the exact behavior the sibling change ships on purpose.
Assert the invariant instead: the expected base set is present, and anything
extra must be inside _RECENTLY_SHIPPED_TOOLSETS (vacuously exact again once
that set empties between releases).
send_video got the 60s read_timeout but send_voice/send_audio/send_photo/
send_document/send_media_group/send_animation upload through the same PTB
request path and hit the same server-side processing wait before the
response arrives. Same class, all sites: they all pass
_MEDIA_SEND_READ_TIMEOUT now. Also drops an unused test helper.
ChatView migrated tip-keyed composer queue/draft entries onto queueSessionKey
whenever the two ids differed. queueSessionKey is route-driven and can flip to
Session B a frame before the store selection leaves Session A, so migrate
re-homed A queue entries onto B and the idle ChatBar auto-drained them into
the wrong chat.
Gate migrate on same-conversation lineage only (tip to root). Also honor
lineage when background queue drain decides selected/busy, so a root queue key
is not treated as idle/offscreen while the compression tip is still working.
The row fell back to the global $repoStatus whenever repoPath was blank,
painting the main pane's branch and ± onto a tile whose cwd hadn't
resolved yet. The fallback bought nothing — the primary computed is keyed
to $currentCwd, which is empty in exactly that case — and cost a rail
showing a tree the session was never in.
A tile and a branched session each live in their own worktree and render
from their own SessionView slice, so neither is the main pane's session.
Pass foreground: false at both call sites.
applyRuntimeInfo unconditionally mirrored a runtime's cwd, branch, model
and usage into the global composer atoms. Every tile create and session
branch called it, so opening a session in another worktree re-pointed the
MAIN pane's coding rail at that tile's repo — and persisted it, so the
wrong workspace cwd survived a restart.
Collect the patch first, then mirror it once behind a `foreground` gate.
Background callers still get the full patch for their own session state;
they just stop publishing into state they don't own.
main now labels each panel row's kebab with the row's name
(menuLabel={profile.name}), so the hardcoded "Actions" default this test
relied on no longer exists. The name alone is ambiguous — the row-select
button carries it too — so match the menu trigger via `expanded`.
Neither side conflicts textually, so this only surfaced once main merged in.
Addresses review on #73013.
1. Manage Profiles used a hand-rolled delete Dialog next to the shared
DeleteProfileDialog in the same folder. That copy missed the active-
profile re-home fix (f764b0400): deleting the profile the gateway is
on stranded it on a dead backend. Switch to the shared dialog, which
owns the deleteProfile call and re-homes to default. Drops
handleConfirmDelete, the deleting state, and the now-unused Dialog*
imports.
2. The name field regressed to a plain Input during the create-dialog
dedup, losing live slugging. Level both shared dialogs up to
SanitizedInput sanitize={slug} so every entry point gets the behavior
Manage Profiles had — the sanitize primitive means callers never
validate-then-reject.
3. Nothing rendered ProfilesView, which is how the drift got in. Add a
behavior test: create dialog exposes SOUL.md, deleting the active
profile re-homes to default, deleting a non-active one does not.
The Manage Profiles page had its own local CreateProfileDialog/
RenameProfileDialog copies that predated the shared dialogs in
create-profile-dialog.tsx / rename-profile-dialog.tsx. The local
create copy lacked the SOUL.md textarea, so New Profile from the
sidebar rail and New Profile from Manage Profiles rendered different
modals.
Delete both local duplicates and reuse the shared self-contained
dialogs (they own the createProfile/renameProfile/updateProfileSoul
calls), so both entry points show the same modal including SOUL.md.
test_generated_script_contains_umask_else_branch asserted on shell
script text ('else', 'umask', '(0666 & ~0', 'chmod') rather than
behavior — a change-detector test per AGENTS.md. The behavioral
test (test_new_file_gets_umask_default_permissions) already
covers the actual behavior end-to-end via real subprocess.
The Docker Build, Test, and Publish workflow fails when
docker/setup-buildx-action can't pull the moby/buildkit:buildx-stable-1
image from Docker Hub. The failure happens during builder bootstrap at
the auth token exchange — a transient network blip (connection reset,
read timeout, rate limiting) that self-resolves on re-run.
Recent failure (run 30449230291, merge job):
read tcp 10.1.0.171:45666->104.18.43.178:443: read: connection reset by peer
This has hit us before and will again — it's the same class of
transient Docker Hub flake that the merge job already retries for
imagetools create. But buildx setup had no retry, so a single network
hiccup killed the entire job (build, publish, or merge) even though
nothing was wrong with the code or the image.
Fix: wrap each of the 3 buildx setup steps (build, publish, merge jobs)
with continue-on-error + a conditional retry step. The maintained action
is preserved as-is — we just give it a second attempt if the first
fails. The action generates a unique builder name per invocation, so the
retry never collides with the failed first attempt. The second attempt
has no continue-on-error, so genuine persistent failures still fail the
job.
The docker/setup-buildx-action maintainer has explicitly said retry
belongs at the workflow level, not inside the action [1], and other
repos use this same continue-on-error pattern for this exact issue [2].
[1] docker/setup-buildx-action#510
[2] joshjhall/containers#688, ethpandaops/eth-client-docker-image-builder#391
The compaction summary's role was selected against the LITERAL
neighbouring messages (compressed[-1] / tail_messages[0]). Mistral-family
chat templates (Devstral, Mistral Small 3.x, Magistral) enforce
user/assistant alternation but exempt the tool flow (tool results and
assistant messages carrying tool_calls) from the check, so a protected
head ending [user, assistant(tool_calls), tool] pinned the summary to
role="user" while the last role the template counts is "user": the
backend rejects the whole request with a Jinja alternation error
(HTTP 500). The summary persists in the stored conversation, every
retry replays the identical poisoned history, and the session is
permanently unrecoverable. Fires on EVERY compaction against a
Mistral-strict backend, captured byte-exact via a tee-proxy in front of
a llama.cpp/llama-swap Devstral deployment.
Fix: compute both neighbour roles through _template_visible_role(),
which skips template-exempt messages. The #52160 (Anthropic user-first)
and #58753 (zero-user-turn) forced-user guards are preserved; their
forced shapes (summary-user followed only by exempt messages) are
alternation-safe. When the visible head ends "assistant" and the
visible tail opens "user", no standalone role can alternate and the
existing merge-into-tail fallback now correctly fires (the literal
logic emitted a standalone user summary there: a second poisoning
shape).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01LGN45sMMbwM8cW9T9ga4ou
