memories/MEMORY.md is the "§"-delimited store written by MemoryStore, not a
markdown document. parse_existing_memory_entries() fell back to
extract_markdown_entries() -- the *source* parser for CLAUDE.md / AGENTS.md --
whenever the destination held no delimiter, which is exactly the case for a
single-entry store or one that was hand-edited or shell-appended. That
extractor skips fenced code blocks, skips table rows, splits a block into one
entry per bullet and reflows paragraphs. The shredded result was then written
straight back over the user's store and reported as "Imported", with no backup
to recover from.
Parse the destination the way MemoryStore._parse_entries does: split on
ENTRY_DELIMITER only, so a store with no delimiter is one intact entry.
extract_markdown_entries() is unchanged and still used on the sources, where
it is correct.
Also restore the safety net the port dropped. The openclaw migration script
this module was ported from calls maybe_backup(destination) before rewriting a
memory store; the port did not. Snapshot the store to <name>.bak.<unix_ts>
(same naming as MemoryStore._backup_drifted_file), refuse to rewrite when the
snapshot fails, and write via temp file + atomic rename so an interrupted
import cannot leave a truncated store and a symlinked MEMORY.md stays a
symlink.
The identical fallback lives in openclaw_to_hermes.py, where it is reached
from migrate_memory() (memories/MEMORY.md and memories/USER.md) and
migrate_daily_memory(); fixed there too.
The old test asserted _warm_gateway_module was fire-and-forget (startup
completes in << SLOW_SECONDS). PR #73291 intentionally reversed this:
the import now runs synchronously before the lifespan yield because
run_in_executor didn't release the GIL on Windows + Python 3.11.
Updated the test to assert startup blocks for >= SLOW_SECONDS.
STT previously had no configuration surface outside hand-editing
config.yaml — no category in the hermes tools picker, no provider
matrix in the GUI capabilities tab, no status line in hermes setup.
- TOOL_CATEGORIES['stt']: 7 provider rows (Local Whisper, Nous
Subscription managed, OpenAI, Groq, xAI, ElevenLabs Scribe,
DeepInfra) with key prompts, badges, and post-setup hooks
- stt_provider marker wired through _write_provider_config,
_configure_provider, _reconfigure_provider, and
_is_provider_active — GUI and CLI share one write path
(apply_provider_selection)
- STT model picker (_configure_stt_model + STT_MODEL_CATALOG) runs
after provider pick: local sizes, Groq whisper family, OpenAI
whisper-1/gpt-4o-*/gpt-transcribe, ElevenLabs scribe (model_id key)
- faster_whisper post-setup hook auto-installs the local backend;
registered in _POST_SETUP_READY
- stt is CONFIG-ONLY (_CONFIG_ONLY_TOOLSETS): it ships no tool
schemas, so it is excluded from the per-platform enable checklist;
the GUI toolset toggle writes stt.enabled instead of
platform_toolsets
- hermes setup shows a Speech-to-Text status line per provider
- Mistral row omitted (mistralai PyPI quarantine), mirroring the
dashboard stt.provider options
Tests: tests/hermes_cli/test_stt_picker.py (20 cases) incl. invariant
checks against agent.transcription_registry builtins and the runtime
OPENAI_MODELS/GROQ_MODELS sets.
Follow-up on the #53205 salvage: replace bare is_file() probes of the
managed (~/.hermes/node[/bin]) and legacy (node_modules/.bin) locations
with shutil.which(..., path=dir) so Windows resolves the executable
.cmd shim instead of the extensionless POSIX script — the same miss
class fixed for _has_agent_browser() in #73932. Also covers the
Windows managed layout where the binary sits in node/ directly.
`hermes acp --setup-browser` installs agent-browser into the Hermes-managed
node prefix (~/.hermes/node/bin/agent-browser), which isn't necessarily on
PATH. doctor only checked PROJECT_ROOT/node_modules and PATH (shutil.which),
so it false-negatived with "agent-browser not installed" even though the
binary was present and runnable. Mirror dep_ensure._has_hermes_agent_browser()
by also checking HERMES_HOME/node/bin and the legacy
HERMES_HOME/node_modules/.bin path, each gated by agent_browser_runnable().
Tested with tests/hermes_cli/test_doctor.py (added positive + not-runnable
cases) and pytest tests/hermes_cli/test_doctor.py -q (66 passed).
Streamed response text carried a 4-space _STREAM_PAD indent and the
final-response Rich Panel used padding=(1, 4), so every line selected
out of the terminal came with leading whitespace. Both now render
flush-left (pad empty, panel padding=(1, 0)); the table-realignment
width budgets were widened to match.
/copy now writes the ORIGINAL message text through native clipboard
tools (pbcopy / PowerShell Set-Clipboard via base64 / wl-copy / xclip /
xsel — same fallback chain as the TUI's writeClipboardText), falling
back to OSC 52 only when no native backend succeeds. This is the
TUI-equivalent answer to soft-wrap mangling: the clipboard gets the raw
text, not the rendered layout.
_has_agent_browser()'s new managed-Node rung calls
shutil.which('agent-browser', path=...); tests that monkeypatch
shutil.which globally with 1-arg lambdas raised TypeError when their
code path reached the browser readiness probe (test_post_setup_gating,
test_setup_model_provider).
Three GUI Capabilities-tab defects reported on Windows:
1. Browser rows stuck on 'Setup required' after a successful setup run.
Root causes, all in the readiness probe (not the installer):
- _has_agent_browser() never searched the Hermes-managed Node dir
(%LOCALAPPDATA%/hermes/node / $HERMES_HOME/node/bin) where the
Windows install lands, and probed node_modules/.bin/agent-browser
as the extensionless POSIX shim, which fails exec on Windows
(WinError 193) — now resolved via PATHEXT-aware shutil.which
against both rungs, mirroring _find_agent_browser().
- Cloud rows (Nous Subscription Browser Use, Browserbase, Browser
Use, Firecrawl) declared post_setup: agent_browser, whose
readiness gate requires a LOCAL Chromium build the cloud never
uses — switched to the cloud-scoped 'browserbase' hook (CLI-only).
- _agent_browser_installed() could read browser_tool's stale cached
'Chromium missing' result from before the install ran in the
spawned post-setup process — cache now dropped before probing so
the pill flips to Ready right after a successful run.
2. No way to tell which backend is active, and clicking a row to read
its details silently rewrote config. Row click now only
expands/collapses; activation is an explicit 'Use this backend'
button, the active row carries an 'Active' pill, and the expanded
active row says 'This is your active backend'.
3. OpenAI TTS showed one model and one voice. The options were always
defined but rendered through a native <datalist>, which filters by
the field's current value — a field already set to a valid option
suggested only itself. Replaced with a real combobox (Input +
dropdown) that lists every option, and voice suggestions now track
the selected model per the OpenAI TTS docs: tts-1/tts-1-hd = 9
voices, gpt-4o-mini-tts = 13 (adds ballad, verse, marin, cedar).
show_status now reads get_nous_auth_status_local(); the test_status.py
mocks still patched the old live-resolve entry point, so the patched
dict was never consumed (CI slice 8/8 red on the salvage PR).
Follow-up widening of the /api/status fix: add get_nous_auth_status_local(),
a refresh-free auth-store snapshot (local invoke-JWT decode only), and use it
on the read-only display surfaces that previously called
get_nous_auth_status() -> resolve_nous_runtime_credentials() -> live OAuth
refresh POST:
- hermes_cli/status.py (hermes status auth-provider panel)
- hermes_cli/doctor.py (hermes doctor auth-provider checks)
- hermes_cli/portal_cli.py (hermes portal status display)
- hermes_cli/web_server.py /api/portal endpoint and the accounts-tab
provider card dispatcher (_resolve_provider_status nous branch)
Action paths (login flows, portal operations needing a live credential)
keep using get_nous_auth_status(). Part of NS-592.
Hosted agents that die uncleanly (kernel OOM kill, SIGKILL, whole-VM
death) leave no trace: shutdown_forensics only covers graceful signals,
gateway-exit-diag.log only covers exit paths that actually run, and the
VM reboot wipes dmesg before anyone can capture it. NS-608 (BlueAtlas
hourly crash cycle, July 12-15) took days of manual log correlation to
classify because nothing recorded 'the previous life ended violently'.
Add gateway/lifecycle_ledger.py — a sentinel state machine persisted to
<HERMES_HOME>/state/gateway.lifecycle.json:
- start_gateway() claims the sentinel (phase=running) right after the
PID-file/runtime-lock claim, and reports any prior life that never
reached an exit path as gateway.previous_unclean_exit in
gateway-exit-diag.log + a WARNING log line.
- Every exit funnel marks the sentinel exited with a reason:
_exit_after_graceful_shutdown (graceful_shutdown), the shutdown
watchdog (shutdown_watchdog), and the loop-liveness watchdog
(loop_liveness_watchdog).
- Ownership-guarded for --replace takeovers: a live matching owner is
never reported dead, and the old life cannot clobber the
replacement's freshly claimed sentinel on its way out.
The 30s loop heartbeat now embeds a cheap /proc memory sample (own RSS,
MemAvailable, swap used) so every unclean-death report carries a
'memory N seconds before death' snapshot; the detector flags
suspected_oom when the last sample shows <64MiB or <5% available.
container-boot.log lines gain prior_exit=clean|unclean|unknown per
profile, stamping unclean container deaths into the volume-persisted
boot log where support can grep for them.
Tests: tests/gateway/test_lifecycle_ledger.py (16 cases) + 4 new
container-boot annotation cases. Existing watchdog/forensics/boot
suites all green; ruff clean.
_install_dependencies now routes through tools.lazy_deps.install_specs
(NS-605); the force-reinstall test still stubbed
hermes_cli.tools_config._pip_install, so its spy list stayed empty
(CI slice 3/8 red on the salvage PR).
Installing a memory provider (Honcho, mem0, hindsight, ...) from the
dashboard Plugins page failed on hosted deployments with a permission
error: the setup endpoint shelled out to
`uv pip install --python sys.executable`, which targets the sealed
read-only venv under /opt/hermes (immutable hosted image, NS-579/#49113).
The correct mechanism already exists: tools/lazy_deps.py redirects
installs to the writable durable target on the data volume
(HERMES_LAZY_INSTALL_TARGET=/opt/data/lazy-packages) when the venv is
sealed (HERMES_DISABLE_LAZY_INSTALLS=1), appends the target to the END
of sys.path (core venv always wins collisions), and constrains shared
deps to core-venv versions. The dashboard installer simply never used
it.
Fix:
- tools/lazy_deps.py: new public install_specs() — installs arbitrary
manifest-declared pip specs through the same environment routing as
ensure(): venv-scoped by default, durable-target on sealed images,
refused with an actionable reason when gated off (config kill switch
or sealed venv without a target — never surfaces raw EROFS/EACCES).
Specs are validated with _spec_is_safe(); post-install it invalidates
import/metadata caches so availability rechecks in the same process
see the new packages without a restart. Never raises.
- hermes_cli/web_server.py: _install_memory_provider_pip_dependencies
now calls install_specs() instead of building its own uv/pip
subprocess. Blocked installs surface the gate reason in the setup
results; the response's status block reflects post-install
availability (stale 'missing deps' state clears immediately).
- hermes_cli/memory_setup.py, plugins/memory/honcho/cli.py,
plugins/memory/mem0/_setup.py: CLI setup wizards routed through
install_specs() too — same sealed-venv failure mode, same fix.
No hosted setup path writes to /opt/hermes anymore; provider discovery
and installation now use the same environment (sys.path activation is
shared with the lazy-install bootstrap in hermes_bootstrap).
Tests:
- tests/tools/test_lazy_deps.py: TestInstallSpecs — gating matrix
(sealed+no-target blocked with immutable-deployment reason, config
kill switch, sealed+target proceeds), spec-safety rejection before
any subprocess, venv-scoped vs --target command display, failure
stderr passthrough, never-raises contract.
- tests/hermes_cli/test_web_server.py: setup endpoint routes pip
through lazy_deps (regression guard asserts no direct 'pip install'
subprocess), blocked-reason surfacing, same-response availability
recheck clears stale missing state.
Fixes NS-605 (Plain T-1111).
speak_text (hermes_cli/voice.py — the TUI/gateway one-shot TTS entry
point) now checks resolve_streaming_provider() first: when the
configured provider has a chunked streamer, the reply is spoken through
the same stream_tts_to_speaker pipeline CLI voice mode uses, so audio
starts on sentence one instead of after whole-file synthesis. No
streamer (edge/piper/etc.) or a streaming failure falls back to the
existing whole-file path unchanged — one dispatcher, zero parallel
streaming implementations.
Refs: #58930
Sibling fix for #65977 — _model_flow_bedrock_api_key used only
get_env_value for AWS_BEARER_TOKEN_BEDROCK, missing pool-backed
keys. Now uses _resolve_api_key_provider_secret like the other
flows.
Sibling of #65254 (main-slot endpoint preservation): the auxiliary scope of
POST /api/model/set dropped the request's base_url/api_key on the floor, so
an aux slot pinned to a custom/local endpoint silently depended on
model.base_url — and broke the moment the main slot switched away and
cleared it. The aux resolver already reads auxiliary.<task>.base_url/api_key
(_resolve_task_provider_model); this persists them.
Desktop side: setAuxiliaryToMain / applyAuxiliaryDraft now carry the
user-defined provider's api_url as base_url, mirroring applyMainModel.
Port from openai/codex#34540 / #34612 ("detach non-interactive
subprocesses from stdin"): internal git invocations that run with nobody
attached — MCP catalog installs, plugin install/update, profile
distribution staging, worktree base fetches, and the desktop review
pane's git/gh backend — could hang on a credential prompt when a remote
is private, misconfigured, or requires auth. git prompts on the
inherited terminal (or via Git Credential Manager on Windows), so the
operation silently waits until its timeout, or forever at sites without
one (mcp_catalog clones have no timeout at all and inherit the parent
terminal).
- Add noninteractive_git_env() to hermes_cli/_subprocess_compat.py:
GIT_TERMINAL_PROMPT=0 + GCM_INTERACTIVE=Never on a copy of the
environment; GIT_ASKPASS/SSH_ASKPASS deliberately preserved so
working non-interactive auth still succeeds.
- Wire it + stdin=DEVNULL into: mcp_catalog._do_git_install (clone/
checkout), plugins_cmd (clone + pull), profile_distribution._git_clone,
web_git._git/_gh (gh also gets GH_PROMPT_DISABLED=1), and cli.py's
worktree base fetch helper.
- Tests: env contract, a real-git E2E against a local 401 Basic-auth
HTTP server proving fail-fast ("terminal prompts disabled") instead
of a hang, and per-call-site plumbing assertions. Sabotage-verified:
removing the env from web_git._git fails the site test.
Hardened-runtime restrictions are enforced even for ad-hoc signatures,
so signing with --options runtime without the allow-jit entitlements
would leave Electron/V8 crashing on launch — strictly worse than the
legacy plain ad-hoc sign. Raise instead, so the fixup falls back to the
legacy path and the bundle always stays launchable.
Local/self-updated macOS builds were finished with a plain
'codesign --force --deep --sign -', leaving a cdhash-only Designated
Requirement and stripping electron-builder's entitlements. Every rebuild
changes the cdhash, so TCC treats the new bundle as different code and
forgets Full Disk Access, Desktop/Downloads/Documents, Accessibility,
Automation, and microphone grants — users re-approve everything after
every update.
Rework the relaunch fixup to sign inside-out (standalone Mach-O
binaries, nested frameworks/helpers, then the main bundle), preserving
the repo's entitlement plists, and pin an identifier-based Designated
Requirement when signing ad-hoc so TCC has a stable identity to persist.
Opt-in desktop.macos_signing_identity names a persistent keychain cert
(self-signed Code Signing cert works — no Apple Developer account) for a
certificate-anchored DR, the strongest form. An intact Developer ID
signature is detected and never clobbered, callers can pass the
publisher-signing decision explicitly so a later dotenv load can't flip
it, and the legacy deep ad-hoc sign remains the last-resort fallback.
Co-authored-by: lewis4x4 <lewis4x4@users.noreply.github.com>
Co-authored-by: natebransc <natebransc@users.noreply.github.com>
Co-authored-by: caseyanthony <caseyanthony@users.noreply.github.com>
Co-authored-by: gvago <gvago@users.noreply.github.com>
Co-authored-by: twe-cloud <twe-cloud@users.noreply.github.com>
The autouse _audio_playback_guard from this PR stubs voice.speak_text
globally — but these tests exercise speak_text itself with their own
playback stubs (no real audio possible). Mark real_audio_playback so
the guard yields; the tests' own monkeypatches keep speakers silent.
/api/audio/transcribe, /api/audio/speak, /api/audio/elevenlabs/voices, and
the /api/audio/speak-stream WebSocket resolved TTS/STT config from the
dashboard's own HERMES_HOME regardless of the active profile, so a
non-default profile's voice settings were silently ignored. Give all four
the same optional profile param as the rest of the dashboard surface,
entering _config_profile_scope (await-safe, config-only — the audio paths
touch no skills globals) inside their worker threads.
Backend half of the desktop fix; completes the renderer-side profileScoped()
threading. Fixes#53441#45506#66012#64057.
Two in-house micro-fixes from issue triage:
- #49883: voice.beep_enabled gates in cli.py and hermes_cli/voice.py used
bool() on the config value, so a quoted YAML string like "false" or
"off" kept beeps on. Route through utils.is_truthy_value.
- #18432: AudioRecorder.start() collapsed OSError from _import_audio into
the 'pip install sounddevice numpy' hint — but OSError means the
PortAudio SHARED LIBRARY is missing, which pip cannot fix. Mirror
detect_audio_environment's system-package hint (libportaudio2 /
brew portaudio / Termux pkg install portaudio) on that path.
Fixes#49883Fixes#18432
Use mark_provider_active_if_unset after dashboard token save, unsuppress
device_code after TTS setup login, and lock default-active plus refresh
active_provider contracts in tests.
Save side-tool OAuth tokens without promoting xai-oauth via active_provider
or model.provider so hermes setup tts login no longer hijacks inference routing.
Existing installs predate the install.sh hermes-acp launcher, and hermes
update never re-runs setup_path, so ACP hosts (Zed, JetBrains, Buzz)
still resolve Hermes as unavailable until a reinstall. _ensure_acp_launcher()
writes the launcher next to an existing hermes command in ~/.local/bin or
/usr/local/bin during hermes update — delegating to the sibling launcher so
it is correct for every install layout. Never follows symlinks (#21454),
skips unwritable dirs, no-op on Windows (venv Scripts is already on PATH).
Docs: add a Buzz Desktop section to the ACP page (en + zh-Hans).
Per-job cron inference pins are now user-owned: the agent-facing cronjob
tool schema no longer exposes model/provider/base_url, and the registered
handler ignores them even if a model hallucinates the old parameters.
Users set pins via the dashboard, hermes cron create/edit --model/--provider,
or jobs.json directly — and once set, a pin sticks until the user changes it.
Existing agent-era pins are grandfathered untouched.
New cron.model / cron.model_provider config keys give the cron fleet its
own default model, independent of the chat model. Fire-time resolution:
per-job pin > cron.model > HERMES_MODEL > model.default. An axis covered
by the explicit cron-fleet default is deliberate routing, not drift, so
the #44585 fail-closed guard skips it — switching your chat model with
/model or hermes model no longer breaks unpinned cron fleets.
- tools/cronjob_tools.py: drop model param from agent schema + handler;
remove now-dead _resolve_model_override
- cron/scheduler.py: cron.model/model_provider resolution + per-axis
drift-guard skip
- cron/jobs.py: snapshot resolution mirrors the new precedence
- hermes_cli/subcommands/cron.py + hermes_cli/cron.py: --model/--provider
on hermes cron create/edit
- hermes_cli/config.py: cron.model / cron.model_provider defaults
- docs: cron.md model-resolution tip rewritten
Responds to hermes-sweeper review on #54717: existing tests only mocked
platform_registry.get(). Add a hermetic fake deferred-loader test that
runs real PlatformRegistry resolution → PluginContext.register_cli_command
→ argparse subparser/choices visibility, without Photon SDK imports.
Closes#54678
`hermes photon ...` could fail with argparse `invalid choice: 'photon'`
even when the bundled Photon platform plugin is present. Photon registers
its top-level CLI command from the platform adapter module via
`ctx.register_cli_command(name="photon", ...)`, but bundled platform
plugins are cheap-registered as *deferred* entries to avoid importing every
gateway SDK during normal startup.
On the unknown-top-level-command slow path, `discover_plugins()` records the
deferred loader but never imports the matching platform module, so the CLI
registration side effect doesn't run and `photon` stays absent from
`_cli_commands` — argparse then rejects it.
Fix: after `discover_plugins()` on that slow path, resolve only the deferred
platform whose name matches the first positional token (via
`platform_registry.get(name)`) before reading `_cli_commands`. This imports
exactly the targeted platform, leaving normal startup cheap (a bare `hermes`
or flags-only invocation has no positional token and touches nothing). The
resolution is best-effort: registry/import failures are logged at debug and
never crash startup.
Added 3 tests in tests/hermes_cli/test_startup_plugin_gating.py: resolves the
matching platform, ignores empty/None command, and swallows registry errors.
Fails without the fix (symbol absent).
- Extract the cmdk filter into exported rankSearchOption and cover it,
SearchableSelect selection/clear/placeholder, and ConfigField
searchable-schema routing with 12 vitest cases (sabotage-verified:
the backend schema test fails on unfixed main).
- Add searchPlaceholder/noResults/systemDefault strings to ja/ar/zh-hant
(zh + en came with the salvaged commits; defineLocale would have
fallen back to English otherwise).
- Add a backend invariant test: timezone ships as a searchable,
clearable select of sorted IANA ids with a UTC fallback.
The cap is shared across CLI, desktop/TUI and the messaging gateway, so the
surface that gets rejected is rarely the one holding the slots. The rejection
read "Hermes is at the active session limit (5/5). Try again when another
session finishes." while every slot was an idle desktop tab, which took
filesystem access to work out.
Name the holders in the message, and show slot usage plus each holder in
`hermes status`. Both are inert when max_concurrent_sessions is unset, which
is the default. The gateway's duplicate copy of the message now reuses the
shared helper.
An open chat window took a session-cap slot at session.create/resume time.
Every desktop tile paint and every background reconnect-resume opens one, so
on a websocket-flappy host they accumulated: five parked desktop tabs filled a
5-slot cap and locked the messaging gateway (which shares the cap) out for
fourteen minutes while running no agents at all.
A slot held that way is invisible everywhere. An unprompted draft has no DB row
and the sidebar filters it out with min_messages=1, so the only way to diagnose
it was reading runtime/active_sessions.json by hand.
Claim on the first turn instead, mirroring the lazy contract
_ensure_session_db_row already uses for the row itself. Capacity now means an
agent can run rather than that a window exists, and anything holding a slot is
something the user can see.
Also reclaim leases whose session skipped teardown. _prune_dead only fires when
the owning pid dies, and a dashboard/serve backend runs for days, so a leaked
lease was held until restart. The owning process reconciles against the leases
it still holds, which is exact and needs no heartbeat write on the turn path.
Replaces the O(N) list_sessions_rich histogram in /api/sessions/stats
with a single GROUP BY query, reducing response time from ~575ms to
<1ms on large databases.
Original PR #48921 by @liuhao1024. Salvage fixes based on review
feedback from teknium1 and @wernerhp:
1. Preserve try/except guard — a DB error still degrades to empty
by_source instead of failing the whole stats response.
2. GROUP BY COALESCE(source, 'cli') — the original GROUP BY source
could emit duplicate 'cli' keys (NULL group + literal 'cli' group)
that the dict comprehension silently dropped.
3. Add exclude_children=True — list_sessions_rich excludes subagent
runs, delegates, and compression continuations by default; the
bare GROUP BY counted all rows, inflating source counts.
Aggregate shape (exclude_children/include_archived/limit params)
adapted from closed duplicate #61120 by @mijanx.
Closes#48914
Co-authored-by: mijanx <mijanx@users.noreply.github.com>
- Replace _load_cron_jobs_for_config_warning with lazy import of
cron.jobs.load_jobs — picks up BOM handling, corruption repair,
and context-local store resolution for free
- Re-add model.name to axis mapping (was dropped during merge);
model.name is a legacy alias for model.default
- Fix grammar: '1 enabled unpinned cron job have' -> 'has'
- Pass user_config to cron_model_drift_guard_enabled in set_config_value
to avoid a redundant load_config() re-read of the file just written
When an operator changes the global model/provider config, warn that
unpinned cron jobs with stored snapshots will fail-closed on their next
run. Adds a cron.model_drift_guard config opt-out (default true) for
fleets that should deliberately track changing global defaults.
Addresses #59031. Original PR #59177 by @doncazper.
Aggregator providers (nous, openrouter) serve dozens of models across many
labs. Add a featured=True enricher to build_models_payload that attaches a
featured_models shortlist to each row: within every lab keep the newest
_FEATURED_PER_LAB models by models.dev release_date, ranked among that row's
own models — never against the current date, so the choice is stable as
models age. Same-date ties fall back to curated list order. Single-lab
providers get an empty list. Derived live from the models.dev catalog already
loaded on this path; no hand-maintained allowlist.
