The Kimi team noticed that traffic from Hermes Coding Plan users
identifies itself as Claude (User-Agent: claude-code/0.1.0) rather
than the actual client. They asked us to update the UA so they can
properly attribute traffic and understand how their services are
accessed — especially important as they open up to more third-party
agents.
Three code paths were sending wrong/attribution-less headers to Kimi:
1. run_agent.py — _apply_client_headers_for_base_url sent
{"User-Agent": "claude-code/0.1.0"} for api.kimi.com. Now sends
the same _AI_GATEWAY_HEADERS set used for Vercel AI Gateway:
HTTP-Referer + X-Title + HermesAgent/{version} User-Agent.
2. agent/anthropic_adapter.py — the Anthropic Messages path for
api.kimi.com/coding sent 'claude-code/0.1.0'. Now sends the same
three-header attribution set.
3. plugins/model-providers/kimi-coding/__init__.py — both kimi and
kimi_cn profiles sent a static 'hermes-agent/1.0' with no
HTTP-Referer or X-Title. Now sends the full three-header set with
a dynamic version, matching the pattern used by the gmi, fireworks,
xai, and ai-gateway provider profiles.
The attribution header set (HTTP-Referer + X-Title + User-Agent) is
the canonical Hermes pattern used for OpenRouter, Vercel AI Gateway,
Fireworks, and other providers that read these headers for traffic
attribution.
Adds the comment-based hotspot convention (no new primitives) across three
guidance surfaces:
- KANBAN_GUIDANCE worker lifecycle: new step 7 — when a file keeps colliding
with siblings or appears in other cards' recent comments, leave a
'hotspot: <path> — <reason>' kanban_comment and repeat it in completion
metadata so the orchestrator can decompose the file first.
- kanban.md (en + zh-Hans): 'Collision hotspots in parallel campaigns'
subsection — the convention, the orchestrator response (2+ flags on one
path => dedicated decomposition card before queuing more work touching
it), and the cross-link to merge-reconciler for conflicts that already
happened.
- merge-reconciler SKILL.md Pitfalls: repeated conflicts on the same file
across rounds are a hotspot signal — flag for decomposition rather than
serially reconciling.
Live-verified: guidance renders once via real import (6152 chars); hotspot
comment round-trips through add_comment -> list_comments -> worker context
on an isolated HERMES_KANBAN_DB; kanban tools, review-surfaces, and
merge-reconciler skill tests green (45 passed).
Design decisions belong to the orchestrator: decide naming schemes,
schemas, file formats, and API shapes before fanning out; never let two
subtree cards decide the same question; stamp every decision into each
dependent card body since workers cannot see sibling context. Mirrored
in the kanban docs (en + zh-Hans) with an exporter/importer worked
example, and bounded KANBAN_GUIDANCE size with an invariant test.
Add a non-terminal "review" status so a worker that finished implementation
can hand off for human review without abusing kanban_block. The old
kanban_block(reason="review-required: ...") convention routed the handoff
through the unblock-loop breaker, so a normal review -> changes -> review
cycle was falsely escalated to triage.
- kanban_db: request_review (running/ready -> review, non-block, emits
review_requested), reopen_review_task (review -> ready/todo, review_reopened),
complete_task accepts review -> done, and a review_dispatch gate (default off,
shared by the dispatcher loop and the gateway health probe).
- kanban_request_review worker tool + `request-review` / `reopen-review` CLI
verbs; tool wired through toolsets, EXPOSED_TOOLS, _POLISHED_TOOLS.
- Gateway notifier wakes the origin subscriber on review_requested and
block_loop_detected; the subscription survives until done/archived, so every
review cycle re-notifies.
- Dashboard PATCH + bulk route the review transitions (request_review /
reopen_review_task) and render the review column.
- goals.py goal-loop and KANBAN_GUIDANCE recognize review as a terminator.
- Docs (reference tables, user guide, AGENTS.md, zh-Hans mirrors) + tests.
needs_input / failed are unchanged: they still route through kanban_block,
still count toward block_recurrences, and still escalate to triage.
Cold CLI time-to-banner was ~1.8s (hermes) / ~2.8s (hermes -w). The banner
path was paying for work the session doesn't need before first input:
- aux availability probes built REAL OpenAI/httpx clients (openai import
~0.3s + SSL context) just to answer check_fns. New aux_probe_mode()
returns a cache-excluded stub; resolution policy unchanged.
- tools/mcp_tool imported the mcp SDK (~260ms, mcp.types pydantic model
construction) at module import even with zero MCP servers configured.
SDK import is now lazy behind _ensure_mcp_sdk(); _MCP_AVAILABLE is a
find_spec probe so every existing gate/test keeps its semantics.
- banner blocked 500ms on the update-check prefetch; now waits 50ms and
defers the warning line to a daemon thread (prints above the prompt).
- banner recomputed get_tool_definitions + skills scan + git state every
launch; now snapshotted to ~/.hermes/cache/banner_snapshot.json keyed on
(config.yaml, .env, checkout rev, toolsets) and replayed on warm launches
with a background refresh. Agent tool list is still computed fresh.
- _resolve_active_context_length probed the Nous portal /models (~200ms
network) per launch; the tool-search gate now prefers the on-disk
context cache when present.
- schema reconciliation re-executed SCHEMA_SQL in a scratch SQLite DB
(~85ms) per SessionDB(); the reference parse is now disk-memoized by
DDL hash (live-DB diffing still runs every startup).
- bundled-skills sync (~120-170ms rglob/hash) moved off the startup path
to a daemon thread; plugin discovery starts in the background and every
synchronous consumer joins via discover_plugins().
- hermes_cli.auth imported httpx eagerly (~30ms); now a lazy proxy that
test monkeypatching still reaches (setattr forwards to the real module).
- fast chat launch: unambiguous 'hermes'/'hermes chat' invocations skip
building all ~40 subcommand parsers (bails to full dispatch on anything
else, incl. container mode).
- -w path: git worktree add runs with checkout.workers=8 (0.6s→0.2s) and
overlaps HermesCLI construction; --skills preload runs in the background
and is folded in at agent init (finalize_preloaded_skills, same
fail-loud contract for fully-unknown skill lists); stale-worktree prune
moved off the banner path.
Warm results (PTY time-to-banner, 5-run): hermes 1.80s → 0.38-0.40s;
hermes -w -s hermes-agent-dev --yolo 2.82s → 0.57-0.69s.
After a partial update (stash restore overwriting providers/base.py with
an older version), the NousProfile singleton was instantiated from a
ProviderProfile class that predates the supports_prompt_cache_key field
(added in f4fb23f3d). Accessing profile.supports_prompt_cache_key raised
AttributeError, crashing every API call with:
'NousProfile' object has no attribute 'supports_prompt_cache_key'
Use getattr(profile, 'supports_prompt_cache_key', False) so a stale
profile degrades to 'no prompt cache key' instead of crashing.
Trim verbose comments in conversation_loop.py and run_agent.py to 2 lines
each. Fix the same bug class in the compression summary path at
chat_completion_helpers.py: remove _thinking_prefill from the explicit
pop tuple and move the generic underscore-key sweep to after
_drop_thinking_only_and_merge_users, so the drop pass can recognize
prefill stubs there too.
- warn (not debug) on final text-turn flush failure: a failure here
reopens the exact #81641 data-loss window with _persist_session as
the only remaining retry, unlike the verify siblings which retry
in-loop; include session id for triage
- trim the flush-site comment to sibling proportion, pointing to the
test module for the full incident narrative
- test: assert _persist_session presence before indexing, so a wiring
change fails with a clean assertion instead of ValueError from max()
A pure-text assistant turn (finish_reason=stop) had no durable write of
its own. Its answer reached the user through the streaming / interim
display path, which is display-only and never touches state.db, and the
first durable write was finalize_turn's _persist_session — after the
loop exits and behind post-turn work that can include micro-compaction's
aux-LLM call.
Anything that ended the process or tore the session down inside that
window lost a reply the user had already been shown. On a remote
(non-loopback) backend the window is easy to hit: WS 1006 closures drive
ws_orphan_reap teardown, and affected sessions ended up with user rows
and zero assistant rows in state.db.
The neighbouring exits of the same loop already close this gap:
* the tool-call exit flushes the assistant(tool_calls) block before
handing control to _execute_tool_calls (#49045)
* the verify-on-stop and pre_verify exits flush final_msg before
appending their nudge (#65919 §7)
Apply that same idiom to the ordinary text exit rather than adding a new
persistence mechanism. The intrinsic _DB_PERSISTED_MARKER dedup makes the
later _persist_session a no-op for this row, so no duplicate rows and no
extra write — the same write, just earlier.
Unlike the tool-call exit, a failed flush must not abort the turn: no
side effect runs after this point and the answer is already produced, so
the failure is logged and _persist_session remains the retry.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Move the legacy ghost-row filter from inside the api_messages loop to
BEFORE repair_message_sequence_with_cursor. Dropping a ghost assistant
row between two user messages creates user→user which the repair can
now fix (previously the repair ran first and missed it).
Promote '[This response was interrupted by a user correction.]' to
module-level _INTERRUPT_SCAFFOLD_MARKER constant — used in both
_apply_active_turn_redirect (checkpoint_parts) and the ghost filter,
so they can never drift.
Update ghost-row test: the two consecutive user messages are now
merged by repair, so check for content as substring.
Sessions already poisoned by the incomplete #73146 else branch still replay
hidden assistant rows whose content is the raw interrupt scaffold. Skip those
rows when building provider messages so old state.db history cannot keep
seeding the echo loop.
The incomplete #73146 else branch still wrote the interrupt checkpoint into
the placeholder assistant row. Mid-tool steers then replayed that scaffold as
the model's own prior reply, which it echoed into a self-replicating ghost
loop. Carry the scaffold only on the user correction's api_content, matching
the assistant-tail branch.
Fix: _LENGTH_CONTINUATION_DROPPED_TOOLS_PREFIX ended with '(' but
_get_continuation_prompt still had f'({tool_list})', producing
'((write_file)' instead of '(write_file)'. Removed the '(' from
the prefix constant — the parenthesis belongs in the interpolation.
Widened: promoted the empty-response nudge (line 6993,
'You just executed tool calls but returned an empty response...')
to _EMPTY_TOOL_RESPONSE_NUDGE constant and added it to the
classifier's recognition set. Same bug class — its
_empty_recovery_synthetic metadata flag doesn't survive SessionDB
projection either.
Test: added parametrize case for the empty-response nudge (7→8 cases).
E2E: verified byte-for-byte string equivalence for all nudge constants.
aed114a69 taught _is_synthetic_compression_user_turn to recognize the
max-iteration nudge as ephemeral runtime scaffolding rather than a human
turn, since its role="user" metadata flag doesn't survive SessionDB
projection and a crash/interrupt mid-turn can persist it durably — becoming
the compaction anchor / auto-focus topic in place of the real task.
conversation_loop.py's retry loop appends several more role="user" rows
with the exact same "ephemeral, metadata-tag-only" shape, none of them
recognized by the classifier:
- The three _get_continuation_prompt variants (length-continuation nudge,
tagged _length_continuation_nudge) — two fixed strings plus a third that
interpolates the dropped-tool-call list.
- _CODEX_INCOMPLETE_NUDGE (codex/responses reasoning-only retry).
- The codex ack-continuation nudge (acknowledgment-only reply re-prompt).
- The dropped-tool-call nudge (tagged _dropped_toolcall_nudge) — persisted
across up to 3 consecutive retries before the finalization pop-loop
strips it; an interrupt/crash before that pop can persist it same as the
max-iteration case.
Promote the previously-inline nudge strings to named module-level constants
in conversation_loop.py (single source of truth for both construction and
recognition), then extend the classifier to recognize all of them — exact
match for the five fixed-content nudges, a stable-prefix check for the
dropped-tool-call continuation variant (its tool list is interpolated so it
can't be exact-matched, same treatment TODO_INJECTION_HEADER already gets).
Imported lazily inside the classifier to avoid a module-load-order cycle —
conversation_loop.py already imports FROM context_compressor.py at call
time for the same reason.
Simplify registration/unregistration to match delegate_tool.py's
hasattr+getattr pattern instead of over-defensive try/except Exception
blocks. Delete inspect.getsource() change-detector test (breaks on
rename, proves nothing the behavioral test doesn't cover).
Net: -73 lines, +35 lines = -38 lines.
A background memory/skill review (agent/background_review.py) forks a
second, complete AIAgent in a daemon thread that deliberately shares the
live agent's own session_id for prompt-cache warmth. Nothing previously
stopped a user's next live turn from starting while that fork was still
mid-conversation, letting both stream against the same session_id and
credentials concurrently. That produced two observable failures:
- Doubled prompt-token accounting on the live turn's own calls (the two
concurrent request/response streams under one session_id confuse the
token-usage bookkeeping), triggering premature context compression.
- A lockup that a normal interrupt could not clear: the review fork is a
fully independent AIAgent with its own _interrupt_requested flag, and
was never added to the parent's _active_children list -- the only list
AIAgent.interrupt() actually walks for cross-agent cancellation -- so a
live-turn Ctrl+C had no propagation path to it at all.
Fix, three files:
1. agent/agent_init.py -- add _background_review_agent /
_background_review_lock tracking state to every AIAgent, mirroring the
existing _active_children pattern.
2. agent/background_review.py -- the review fork now registers itself on
the parent's _active_children right after construction (reusing the
same list/lock interrupt() already fans out to for real subagent
delegation), and unregisters on every exit path (success, the
tool-whitelist finally, and the outer exception safety-net). All
registration is defensive (getattr/try-except) so an AIAgent built
without going through agent_init.py's setup degrades to "no
cross-turn cancellation" instead of aborting the whole review.
3. agent/conversation_loop.py -- at the very start of every
run_conversation() turn, if a prior background review is still
in-flight, it is now proactively cancelled via interrupt() before the
live turn proceeds -- fire-and-forget, non-blocking, adds no latency.
Adds 3 regression tests to tests/run_agent/test_background_review.py,
confirmed to fail against the pre-fix code via a scripted revert.
Verified: ruff clean on all touched files; 66/66 background-review and
interrupt-propagation tests pass; 256/256 across turn_finalizer +
run_agent regression suites; no fork-only symbols in the diff.
Follow-ups from review of #82049:
- extract append_user_instruction() into agent/skill_commands so the
stable-prefix construction cannot drift between the skill and cron
builders (the registered prefix must stay a byte-prefix of the built
message); cron no longer imports the private _SINGLE_SKILL_INSTRUCTION
- add the startswith guard to the skill builder registration site,
matching the stronger cron guard
- rename _MAX_BYTES to _MAX_CHARS (sum(map(len, ...)) counts characters,
not bytes) and correct the comment
- collapse find_stable_prefix's two-lock dance into a single critical
section (scan is <=32 short-circuiting startswith calls, measured
2-4us; drops the snapshot copy and the TOCTOU re-check)
- document the split-shape lifetime (marked-endpoint window) in the
module docstring
- add a contract test for the helper's byte-prefix invariant
(mutation-checked)
Follow-up review of the builder-declared cache boundary (#81867) found three
ways the split could silently stop paying off, or keep paying more than it
should, on a long-lived gateway process.
Flattening no longer consults the registry. `strip_anthropic_cache_control`
matched the decorated split by looking the first block up in the prefix
registry, so a mid-turn failover that re-decorates a request built many
messages earlier (#72626) would fail to flatten once _MAX_ENTRIES newer
scaffolds had been registered in between, and would hand the next provider
the two-part shape instead of the canonical string. The split is now matched
by its shape: a marker on the *first* part of a user message is something no
other decoration produces (list content otherwise gets its marker on the last
part, and the two-part [static, volatile] split is role-gated to system), so
the ""-join stays provably byte-exact without any process state. This drops
`is_registered_stable_prefix` and one lock acquisition per stripped message.
Lookups now refresh LRU position. A scaffold fired every minute by cron could
be evicted by a burst of one-off skill invocations while still being the
hottest prefix in the process, silently reverting it to whole-message caching.
Registration now also evicts by total retained bytes (4 MiB). Entries hold
whole expanded skill bodies, so a 32-entry cap alone does not bound memory.
The newest entry is always kept, so a single oversized scaffold still gets a
boundary instead of disabling the split.
Tests: eviction-then-failover round-trip, LRU refresh on hit, byte-cap
eviction, and oversized-single-entry survival.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Webhook/cron skill invocations concatenate a large static scaffold
(activation note + expanded skill body) with a small volatile tail
(ticket payload, timestamps) into one user string, and the Anthropic
cache planner marked that whole string as a single atomic block — so a
few changed tail bytes forced a full cache rewrite on every invocation.
Instead of re-parsing scaffold marker strings out of the message at
request time (fragile when a payload or skill body quotes the marker),
the builders now register the exact stable-prefix bytes in a small
process-local LRU registry at construction time. The cache planner
splits a registered user string into [marked stable prefix, unmarked
volatile tail] request-locally; canonical session history stays a plain
string, and the failover stripper flattens the split back byte-exactly
via an O(1) registry lookup. Unregistered messages keep the existing
whole-message policy.
Covers the single-skill builder (webhook + slash command + TUI) and the
cron job prompt assembler (multi-skill, bundles, skipped-skill notice),
with registration guarded against injection-scanner sanitization.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Folds the model-switch fix in with the untitled retry. They answer
different halves and each is wrong alone: counting alone left a session
that merely opened with machinery nameless forever, because nothing
reconsidered it, and the stored title alone would never title at all on a
store too old to report one. Skip only when both agree — past the opening
turn, and already named.
Counting a turn now judges a multimodal one on its text, so "here's a
screenshot, fix the login" counts as the question it is rather than
reading as machinery and undercounting the conversation.
Co-authored-by: yy28 <yy28@vip.sina.com>
Switching models before sending the first real message titled the session
"[System: The active model for this chat has…" instead of the user's actual
question.
`_append_model_switch_marker` persists its notice with `role="user"` because
strict OpenAI-compatible providers reject a system message that is not first
(#48338). Titling had no way to tell that apart from a genuine opening turn,
which caused two distinct failures:
1. `_MACHINE_PREFIXES` did not cover the marker. Its `[System: ` prefix
matches none of `[CONTEXT COMPACTION`, `[Runtime note:`, or `[SYSTEM]`
(different case, no closing bracket), so `is_titleable_user_message()`
returned True and the marker was formatted into the title.
2. `maybe_auto_title()` counted the marker as a user message. With the marker
present, the first real question arrived at `user_msg_count == 2` and the
`> 1` guard returned early, so the session was never titled at all and its
`title` column stayed NULL. Fixing only (1) would therefore have traded a
wrong title for a permanently missing one.
Add the marker prefix to `_MACHINE_PREFIXES` (kept in sync with
`tui_gateway.server._MODEL_SWITCH_MARKER_PREFIX`) and count only titleable
user messages when detecting the opening turn.
The guard stays narrow: ordinary user text that happens to start with
"[System:" still titles normally.
Adds 6 regression tests, verified to fail without the fix.
Two lookalike gaps found auditing the titler.
_MACHINE_PREFIXES missed the compressor's legacy summary opener and the
"[System note:" injections, so a compacted or resumed session could be
named after the note that carried it. Take the summary prefix from the
compressor that emits it rather than keeping a fourth local copy.
The fast-model exclude list covered embedders but not the other non-chat
siblings a provider names after its chat model — "gpt-4o-mini-tts"
satisfies the "-mini" rung and cannot answer a prompt.
The fast-model picker reads /v1/models to find the small model a provider
currently serves, and it asked anonymously. Most of those endpoints need a
key, so the fetch 401'd and the empty result read as "this provider has no
small model" — the picker fell back to its curated list and never noticed.
Worse, a failed fetch cached its empty result forever, so one bad moment
during startup disabled live model discovery for the life of the process,
and the processes that read this run for weeks. Give the failure an expiry
and pass the provider's credentials.
The bare family rungs (-mini, -flash, haiku) also picked whichever id
sorted first, which is the oldest generation a provider still serves:
gpt-3.5-mini over gpt-5.4-mini, claude-3-haiku over claude-haiku-4.5.
Compare the digit runs as numbers so the rung meant to keep us current
does.
Titling is two-stage — a slice of the user's own words lands inline, the
model's version replaces it a second later — and the platform rename lanes
fired on both. That is two rate-limited calls to reach one name, and
Discord allows two channel renames per ten minutes, so the throwaway could
be the one that survived. The callback now carries which stage it is, and
the lanes take the model's.
The relay lane also asked where the reply landed at title time, which is
before the model has answered: it polled the send-result cache for ten
seconds and read the timeout as "never auto-threaded", so any turn with
tool calls in it silently kept its raw thread name. Wait on the send
itself instead — the adapter already owns that cache, so it can say when a
reply arrives and, just as usefully, that one arrived carrying nothing.
The turn prologue titles every session, and it is shared by every agent —
including the ones no person is reading. A cron job already names its own
session after the job in its finally block, so the titler spent a side-LLM
call per fire to write the delivery scaffolding over it for the length of
the run. A delegated child's session is hidden from every picker, so a
batch at max_concurrent_children paid N title calls for N names nobody
opens.
Both are the same class of run that already sets skip_memory to stay off
the auxiliary path, so keep the titler off it too.
_PROVIDER_PREFIXES was a hand-maintained frozenset, so providers that ship
as plugins (bundled like fireworks, or user plugins under
$HERMES_HOME/plugins/model-providers/) were never recognised as
provider: prefixes in model strings, and metadata/context-window lookups
received the unstripped string. Mirror the _URL_TO_PROVIDER auto-extend
that already sits below it: add each registered profile's name and
aliases after discovery. The _OLLAMA_TAG_PATTERN guard keeps model:tag
strings intact.
Fixes#66106
The native Relay pipeline binds its Futures to the event loop that
entered run_in_session_async. While a managed tool callback executes,
that loop is blocked until the callback returns — so a nested managed
relay call made from inside the callback (vision_analyze's auxiliary
LLM call on a worker-thread loop) awaits a Future that can never
complete: 'RuntimeError: Future attached to a different loop', or a
deadlock, plus 'Event loop is closed' at shutdown when the orphaned
future completes late. (#77244)
Fix: managed_callback_guard, a ContextVar depth marker set around every
Hermes callback the relay adapters hand to the native pipeline
(relay_tools.execute invoke, relay_llm execute/execute_async invoke,
ManagedLlmStream run_callback). resolve_execution_context returns the
no-relay triple while the marker is set, so nested calls run unmanaged.
The marker propagates through contextvars.copy_context() into the
worker threads tools use for their internal async work.
Top-level turn LLM calls and tool wraps stay fully managed — verified
live: vision_analyze works under active shared metrics while the main
turn still records managed llm.execute events.
Alternative fixes considered and rejected: removing
retain_managed_execution (kills the shared-metrics managed pipeline)
and gating on main-thread identity (managed tool wraps legitimately
run on the run_agent thread, so that gate disables relay everywhere).
The Anthropic SDK's streaming accumulator builds ParsedMessage snapshots
whose ParsedTextBlock content doesn't match the generic union pydantic
expects, so model_dump() on stream events (message_stop) emits
PydanticSerializationUnexpectedValue UserWarnings straight into the
user's CLI output mid-response.
Pass warnings=False at every helper that dumps arbitrary SDK models
(relay_llm/_jsonable, relay_tools/_jsonable, anthropic_adapter
_to_plain_data, run_agent _hook_jsonable, chat_completion_helpers
extra_content/reasoning_details sites, chat_completions transport),
with a TypeError fallback for duck-typed model_dump implementations.
Adds regression tests including a precondition test that proves the
fixture still trips the warning without suppression.
Per-turn .env adoption could rewrite agent.api_key while leaving
_credential_pool_entry_id on a previously rotated fallback. The next 429
then marked the healthy fallback exhausted via credential_id precedence
(#79156).
- Sync pool entry id after a successful env credential refresh
- First look does not stomp a pool-rotated key with the env primary
- mark_exhausted_and_rotate prefers api_key_hint when it disagrees with
credential_id
Fixes#79156
_resource_attributes() in otlp_exporter.py built its own hardcoded
resource dict (service.name/instance.id/telemetry.scope only) instead
of reusing gateway_health_export.py's _runtime_resource_attributes(),
which already applies the resource_attributes allowlist from config.
Result: operator-configured attributes like deployment.environment.name
reached metrics and diagnostic logs but never spans.
Span resource building now delegates to the same
_runtime_resource_attributes() helper metrics/logs already use,
removing the duplicate implementation instead of patching it in place.
Generic thinking fields (reasoning / reasoning_content + the
reasoning_details text charge) are replayed for at most the NEWEST
assistant turn on every transport: Anthropic strips all-but-newest at
convert time, Bedrock Converse never replays thinking, and strict
chat-completions providers reject or one-space-pad the field. The tail
budget walks charged them on every message anyway, spending 19-24% of
the budget (per the issue's 1,025-message measurement) on bytes that
provably never reach the wire — so the tail cut landed early and each
compaction discarded more real transcript than configured.
_estimate_msg_budget_tokens now partitions the replay keys:
* _ALWAYS_REPLAYED_BUDGET_KEYS (codex_reasoning_items,
codex_message_items) — charged unconditionally. These ride the wire
on every retained turn (#55572), and codex_reasoning_items now also
carries native server-side compaction checkpoints (#81747).
* _NEWEST_TURN_ONLY_BUDGET_KEYS (reasoning, reasoning_content) + the
reasoning_details text charge — charged only for the newest assistant
turn via charge_stale_thinking, resolved by the three budget walks
(tail cut, raw-budget re-walk, proactive-prune boundary).
Default stays the conservative full charge for callers without
turn-position context. A partition invariant test pins that any future
_REPLAY_BUDGET_KEYS entry must be classified into exactly one class.
Direction credit: #73669 (@x7peeps) and #73730 (@webtecnica) both
attacked this; the keep_open reviews asked for provider/API-mode-aware
accounting that keeps Codex carriers charged — this implements that
shape.
The HUD-mode note tells the model that an unqualified "this" means the app
behind the strip. It says nothing about the app that was behind it a minute
ago, and the user drags the strip from app to app mid-thought: parked over
Spotify, "pause that and play X here" is one request spanning two apps, and
only the second half has a window under it.
Those earlier windows are already in context as read_window_below results, so
the note only has to say they still count. Without it the latest window reads
as the only one and half the request is silently dropped.
No new tool names, so the existing gating tests cover it unchanged.
Titling fired on the first response, so a session sat unnamed for the whole
opening turn - p50 151s, p90 1212s across real sessions, because a turn is
tool calls, not one round-trip. A turn that failed or was interrupted never
got a title at all. Four surfaces each carried their own copy of the call.
Move it into the shared turn prologue and split it in two: a deterministic
title derived from the user's opening message, written inline before the
model runs, then one small-model call that upgrades it. The response is
constrained to a JSON object so there is no preamble to strip, and control
wrappers are stripped rather than refused, so a slash command titles as
what the user asked for instead of the command itself.
Titling ran on the user's main chat model, so a five-word title was billed
to a frontier reasoning model and inherited its latency. Pinning a cheap
model id instead just moves the problem: the hardcoded default was already
dead upstream and every call paid a 404 before the retry net caught it.
Match model FAMILIES against the provider's live /v1/models catalog,
preferring rolling '-latest' aliases where a provider publishes them, and
order the families by measured latency. Nothing to bump when a provider
ships a new mini/flash/haiku. Opt-in per task, so compression, vision, and
search keep 'auto means my chat model'.
A session title had no notion of who set it, so two bugs followed. An
auto-generated title could clobber a name the user typed, and every
compression rotation renumbered the conversation it forked - one piece of
work reaching 'Smallville Map Architecture Plan #10' in the sidebar.
Titles now carry a source (derived < llm < user) enforced by one
compare-and-swap, so an automatic write can only ever replace a title of
strictly lower authority. Compression carries the name across unchanged.
Legacy NULL rows rank as user, so auto-titling only fills genuinely
empty titles on existing data.
The generated-media surface previously had almost no upscaler coverage:
only fal-ai/flux-2-pro chained Clarity Upscaler (hardcoded catalog
default), every other image model returned ~1MP output with no high-res
path, and video had no upscaler at all. Krea's API treats the enhancer
as a standard second pass; this brings the same shape to Hermes.
- image_generate: new optional 'upscale' boolean in the tool schema.
Explicit true chains the backend upscaler on ANY model (including
edits); explicit false disables flux-2-pro's automatic default;
omitted keeps per-model catalog behavior. Response now reports
'upscaled' so the agent knows which resolution it got.
- FAL image path: explicit flag overrides the catalog 'upscale' default
(Clarity Upscaler, 2x). Failure falls back to the native image.
- Krea plugin: upscale=true chains Krea Enhance
(/generate/enhance/krea/enhance, 2x, prompt-guided) through the same
BYO/managed base URL + auth as generation, with a best-effort poll
loop that never fails a successful generation.
- video_generate: new optional 'upscale' boolean; FAL video plugin
chains ByteDance SeedVR2 (fal-ai/seedvr/upscale/video, 2x factor
mode). Providers without upscalers ignore the kwarg per the ABC
contract (documented in both ABCs).
Validation: targeted suites green (123 tests across 6 files, including
new coverage for override-wins/default-kept/failure-fallback on all
three paths); live E2E on direct FAL verified both chains end-to-end
(klein 9b + Clarity upscaled image; pixverse-v6 1s 360p + SeedVR2
upscaled video).
Two corrections on top of the #71077 base (the whole bug class):
1. Turn boundary = last USER message, not last assistant message. A Codex
turn spans several assistant messages (assistant+tool_calls -> tool ->
... -> final assistant) whose reasoning items must replay together; the
last-assistant boundary would strip reasoning mid-chain from the active
turn (the gap flagged in PR #71077 review).
2. type="compaction" checkpoints (native server-side compaction, PR #81747)
are exempt: they carry already-pruned history, not per-turn reasoning.
Pruning filters items instead of popping the sidecar key.
Sibling site fixed in the same class: the Codex incomplete-continuation
dedup path blind-overwrote codex_reasoning_items on visually-duplicate
interim messages, which would drop the only copy of a checkpoint captured
on the earlier response. Extracted merge_interim_reasoning_items() into
agent/native_compaction.py; newer reasoning wins, prior checkpoints are
preserved unless the newer payload carries its own.
In HUD mode Hermes is a strip over the app the user is actually working
in, so "what's under you?" or "look up the weather" is almost always
about that app — but the agent had no way to know it was floating, and
answered from its own browser and panes instead.
The desktop tags a HUD submit with `surface: 'hud'` and the gateway turns
that into a per-turn note pointing at read_window_below, and at carrying
the work out in the app underneath. It rides the model-bound message
beside the reaction and speech-interrupted notes rather than the system
prompt: one session can be driven from the app window on one turn and the
HUD on the next, and the system prompt has to stay byte-stable.
Every tool the note names is checked against the agent's own schema
first, so a session without computer_use or read_window_below is never
pointed at a tool it cannot call.
Complements the cherry-picked contributor fixes and closes out the
remaining sites of the 'missing explicit encoding' bug class, which is
now permanently gated by ruff PLW1514 (enabled repo-wide in
pyproject.toml and enforced by the blocking `ruff check .` step in
.github/workflows/lint.yml):
- tools/memory_tool.py: read MEMORY.md/USER.md via utf-8-sig so a
Notepad BOM never glues U+FEFF onto the first entry (issue #10878,
PR #10888 by @easyvibecoding — strict-decode contract of
_read_raw_checked preserved rather than errors="replace", so
undecodable files still refuse read-modify-write instead of being
lossily rewritten). Regression tests included.
- tools/skills_tool.py: SKILL.md and skill file reads pinned to
utf-8-sig + errors="replace" — deterministic across platforms instead
of the locale fallback proposed in PR #51701 (superseded: falling back
to cp1252/GBK makes the same skill render differently per host); .env
reader aligned with the canonical utf-8-sig dialect in hermes_cli/config.py.
- agent/shell_hooks.py, hermes_cli/main.py, gateway/slash_commands.py:
explicit utf-8 on the remaining fdopen/open text-mode sites flagged by
the AlexFucuson9 sweep series (#56033#56940#65565#66782#66791).
Co-authored-by: easyvibecoding <easyvibecoding@users.noreply.github.com>
Co-authored-by: AlexFucuson9 <AlexFucuson9@users.noreply.github.com>
Co-authored-by: flyingdoubleg <wangzhe00zju@gmail.com>
Co-authored-by: LeonSGP43 <cine.dreamer.one@gmail.com>
_remove_env_source() decides whether a credential var lives in ~/.hermes/.env
or the shell by scanning the .env with env_path.read_text(errors="replace") —
no encoding. read_text() with no encoding falls back to the system locale
(cp1252/GBK on Windows) and never strips a BOM.
The canonical .env readers in hermes_cli/config.py all use
encoding="utf-8-sig" precisely because 'users may edit .env in Notepad which
adds one' (a BOM), and doctor.py documents that .env is written as UTF-8
everywhere. This sibling reader diverged: on a Notepad-edited .env the BOM
prefixes the first line, so line.strip().startswith(f"{env_var}=") is False
for the first variable — the detector reports a .env-backed key as a phantom
shell export and prints a misleading 'still set in your shell environment'
hint on .
Match the canonical reader (utf-8-sig + errors=replace). Adds a regression
test with a BOM'd .env.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
kshitij
Teknium
Jakub Wolniewicz
Nikita Barkov
Laith Weinberger
Jefftree
joaomarcos
miura
HexLab98
pierrenode
Adam Durham
brooklyn!
yy28
Oliver Mee
686f6c61
webtecnica
solyanviktor-star