The streaming translator in agent/gemini_cloudcode_adapter.py keyed OpenAI
tool-call indices by function name, so when the model emitted multiple
parallel functionCall parts with the same name in a single turn (e.g.
three read_file calls in one response), they all collapsed onto index 0.
Downstream aggregators that key chunks by index would overwrite or drop
all but the first call.
Replace the name-keyed dict with a per-stream counter that persists across
SSE events. Each functionCall part now gets a fresh, unique index,
matching the non-streaming path which already uses enumerate(parts).
Add TestTranslateStreamEvent covering parallel-same-name calls, index
persistence across events, and finish-reason promotion to tool_calls.
When the model omits old_text on memory replace/remove, the tool preview
rendered as '~memory: ""' / '-memory: ""', which obscured what went wrong.
Render '<missing old_text>' in that case so the failure mode is legible
in the activity feed.
Narrow salvage from #12456 / #12831 — only the display-layer fix, not the
schema/API changes.
Third-party gateways that speak the native Anthropic protocol (MiniMax,
Zhipu GLM, Alibaba DashScope, Kimi, LiteLLM proxies) now work end-to-end
with the same feature set as direct api.anthropic.com callers. Synthesizes
eight stale community PRs into one consolidated change.
Five fixes:
- URL detection: consolidate three inline `endswith("/anthropic")`
checks in runtime_provider.py into the shared _detect_api_mode_for_url
helper. Third-party /anthropic endpoints now auto-resolve to
api_mode=anthropic_messages via one code path instead of three.
- OAuth leak-guard: all five sites that assign `_is_anthropic_oauth`
(__init__, switch_model, _try_refresh_anthropic_client_credentials,
_swap_credential, _try_activate_fallback) now gate on
`provider == "anthropic"` so a stale ANTHROPIC_TOKEN never trips
Claude-Code identity injection on third-party endpoints. Previously
only 2 of 5 sites were guarded.
- Prompt caching: new method `_anthropic_prompt_cache_policy()` returns
`(should_cache, use_native_layout)` per endpoint. Replaces three
inline conditions and the `native_anthropic=(api_mode=='anthropic_messages')`
call-site flag. Native Anthropic and third-party Anthropic gateways
both get the native cache_control layout; OpenRouter gets envelope
layout. Layout is persisted in `_primary_runtime` so fallback
restoration preserves the per-endpoint choice.
- Auxiliary client: `_try_custom_endpoint` honors
`api_mode=anthropic_messages` and builds `AnthropicAuxiliaryClient`
instead of silently downgrading to an OpenAI-wire client. Degrades
gracefully to OpenAI-wire when the anthropic SDK isn't installed.
- Config hygiene: `_update_config_for_provider` (hermes_cli/auth.py)
clears stale `api_key`/`api_mode` when switching to a built-in
provider, so a previous MiniMax custom endpoint's credentials can't
leak into a later OpenRouter session.
- Truncation continuation: length-continuation and tool-call-truncation
retry now cover `anthropic_messages` in addition to `chat_completions`
and `bedrock_converse`. Reuses the existing `_build_assistant_message`
path via `normalize_anthropic_response()` so the interim message
shape is byte-identical to the non-truncated path.
Tests: 6 new files, 42 test cases. Targeted run + tests/run_agent,
tests/agent, tests/hermes_cli all pass (4554 passed).
Synthesized from (credits preserved via Co-authored-by trailers):
#7410 @nocoo — URL detection helper
#7393 @keyuyuan — OAuth 5-site guard
#7367 @n-WN — OAuth guard (narrower cousin, kept comment)
#8636 @sgaofen — caching helper + native-vs-proxy layout split
#10954 @Only-Code-A — caching on anthropic_messages+Claude
#7648 @zhongyueming1121 — aux client anthropic_messages branch
#6096 @hansnow — /model switch clears stale api_mode
#9691 @TroyMitchell911 — anthropic_messages truncation continuation
Closes: #7366, #8294 (third-party Anthropic identity + caching).
Supersedes: #7410, #7367, #7393, #8636, #10954, #7648, #6096, #9691.
Rejects: #9621 (OpenAI-wire caching with incomplete blocklist — risky),
#7242 (superseded by #9691, stale branch),
#8321 (targets smart_model_routing which was removed in #12732).
Co-authored-by: nocoo <nocoo@users.noreply.github.com>
Co-authored-by: Keyu Yuan <leoyuan0099@gmail.com>
Co-authored-by: Zoee <30841158+n-WN@users.noreply.github.com>
Co-authored-by: sgaofen <135070653+sgaofen@users.noreply.github.com>
Co-authored-by: Only-Code-A <bxzt2006@163.com>
Co-authored-by: zhongyueming <mygamez@163.com>
Co-authored-by: Xiaohan Li <hansnow@users.noreply.github.com>
Co-authored-by: Troy Mitchell <i@troy-y.org>
Follow-up to #12144. That PR standardized the kimi-k2.* temperature lock
against the Coding Plan endpoint (api.kimi.com/coding/v1) docs, where
non-thinking models require 0.6. Verified empirically against Moonshot
(April 2026) that the public chat endpoint (api.moonshot.ai/v1) has a
different contract for kimi-k2.5: it only accepts temperature=1, and rejects
0.6 with:
HTTP 400 "invalid temperature: only 1 is allowed for this model"
Users hit the public endpoint when KIMI_API_KEY is a legacy sk-* key (the
sk-kimi-* prefix routes to Coding Plan — see hermes_cli/auth.py). So for
Coding Plan subscribers the fix from #12144 is correct, but for public-API
users it reintroduces the exact 400 reported in #9125.
Reproduction on api.moonshot.ai/v1 + kimi-k2.5:
temperature=1.0 → 200 OK
temperature=0.6 → 400 "only 1 is allowed" ← #12144 default
temperature=None → 200 OK
Other kimi-k2.* models are unaffected empirically — turbo-preview accepts
0.6 and thinking-turbo accepts 1.0 on both endpoints — so only kimi-k2.5
diverges.
Fix: thread the client's actual base_url through _build_call_kwargs (the
parameter already existed but callers passed config-level resolved_base_url;
for auto-detected routes that was often empty). _fixed_temperature_for_model
now checks api.moonshot.ai first via an explicit _KIMI_PUBLIC_API_OVERRIDES
map, then falls back to the Coding Plan defaults. Tests parametrize over
endpoint + model to lock both contracts.
Closes#9125.
Smart model routing (auto-routing short/simple turns to a cheap model
across providers) was opt-in and disabled by default. This removes the
feature wholesale: the routing module, its config keys, docs, tests, and
the orchestration scaffolding it required in cli.py / gateway/run.py /
cron/scheduler.py.
The /fast (Priority Processing / Anthropic fast mode) feature kept its
hooks into _resolve_turn_agent_config — those still build a route dict
and attach request_overrides when the model supports it; the route now
just always uses the session's primary model/provider rather than
running prompts through choose_cheap_model_route() first.
Also removed:
- DEFAULT_CONFIG['smart_model_routing'] block and matching commented-out
example sections in hermes_cli/config.py and cli-config.yaml.example
- _load_smart_model_routing() / self._smart_model_routing on GatewayRunner
- self._smart_model_routing / self._active_agent_route_signature on
HermesCLI (signature kept; just no longer initialised through the
smart-routing pipeline)
- route_label parameter on HermesCLI._init_agent (only set by smart
routing; never read elsewhere)
- 'Smart Model Routing' section in website/docs/integrations/providers.md
- tip in hermes_cli/tips.py
- entries in hermes_cli/dump.py + hermes_cli/web_server.py
- row in skills/autonomous-ai-agents/hermes-agent/SKILL.md
Tests:
- Deleted tests/agent/test_smart_model_routing.py
- Rewrote tests/agent/test_credential_pool_routing.py to target the
simplified _resolve_turn_agent_config directly (preserves credential
pool propagation + 429 rotation coverage)
- Dropped 'cheap model' test from test_cli_provider_resolution.py
- Dropped resolve_turn_route patches from cli + gateway test_fast_command
— they now exercise the real method end-to-end
- Removed _smart_model_routing stub assignments from gateway/cron test
helpers
Targeted suites: 74/74 in the directly affected test files;
tests/agent + tests/cron + tests/cli pass except 5 failures that
already exist on main (cron silent-delivery + alias quick-command).
- only use the native adapter for the canonical Gemini native endpoint
- keep custom and /openai base URLs on the OpenAI-compatible path
- preserve Hermes keepalive transport injection for native Gemini clients
- stabilize streaming tool-call replay across repeated SSE events
- add follow-up tests for base_url precedence, async streaming, and duplicate tool-call chunks
- add a native Gemini adapter over generateContent/streamGenerateContent
- switch the built-in gemini provider off the OpenAI-compatible endpoint
- preserve thought signatures and native functionResponse replay
- route auxiliary Gemini clients through the same adapter
- add focused unit coverage plus native-provider integration checks
Imperative memory entries ('Always respond concisely', 'Run tests with
pytest -n 4') get re-read as directives in future sessions, causing
repeated work or overriding the user's current request. Add a short
phrasing guideline to MEMORY_GUIDANCE so the model writes declarative
facts instead ('User prefers concise responses', 'Project uses pytest
with xdist').
Credit: observation from @Mariandipietra on X.
The cherry-picked salvage (admin28980's commit) added codex headers only on the
primary chat client path, with two inaccuracies:
- originator was 'hermes-agent' — Cloudflare whitelists codex_cli_rs,
codex_vscode, codex_sdk_ts, and Codex* prefixes. 'hermes-agent' isn't on
the list, so the header had no mitigating effect on the 403 (the
account-id header alone may have been carrying the fix).
- account-id header was 'ChatGPT-Account-Id' — upstream codex-rs auth.rs
uses canonical 'ChatGPT-Account-ID' (PascalCase, trailing -ID).
Also, the auxiliary client (_try_codex + resolve_provider_client raw_codex
branch) constructs OpenAI clients against the same chatgpt.com endpoint with
no default headers at all — so compression, title generation, vision, session
search, and web_extract all still 403 from VPS IPs.
Consolidate the header set into _codex_cloudflare_headers() in
agent/auxiliary_client.py (natural home next to _read_codex_access_token and
the existing JWT decode logic) and call it from all four insertion points:
- run_agent.py: AIAgent.__init__ (initial construction)
- run_agent.py: _apply_client_headers_for_base_url (credential rotation)
- agent/auxiliary_client.py: _try_codex (aux client)
- agent/auxiliary_client.py: resolve_provider_client raw_codex branch
Net: -36/+55 lines, -25 lines of duplicated inline JWT decode replaced by a
single helper. User-Agent switched to 'codex_cli_rs/0.0.0 (Hermes Agent)' to
match the codex-rs shape while keeping product attribution.
Tests in tests/agent/test_codex_cloudflare_headers.py cover:
- originator value, User-Agent shape, canonical header casing
- account-ID extraction from a real JWT fixture
- graceful handling of malformed / non-string / claim-missing tokens
- wiring at all four insertion points (primary init, rotation, both aux paths)
- non-chatgpt base URLs (openrouter) do NOT get codex headers
- switching away from chatgpt.com drops the headers
Follow-up on top of mvanhorn's cherry-picked commit. Original PR only
wired request_timeout_seconds into the explicit-creds OpenAI branch at
run_agent.py init; router-based implicit auth, native Anthropic, and the
fallback chain were still hardcoded to SDK defaults.
- agent/anthropic_adapter.py: build_anthropic_client() accepts an optional
timeout kwarg (default 900s preserved when unset/invalid).
- run_agent.py: resolve per-provider/per-model timeout once at init; apply
to Anthropic native init + post-refresh rebuild + stale/interrupt
rebuilds + switch_model + _restore_primary_runtime + the OpenAI
implicit-auth path + _try_activate_fallback (with immediate client
rebuild so the first fallback request carries the configured timeout).
- tests: cover anthropic adapter kwarg honoring; widen mock signatures
to accept the new timeout kwarg.
- docs/example: clarify that the knob now applies to every transport,
the fallback chain, and rebuilds after credential rotation.
Context compaction summaries were always produced in English regardless
of the conversation language, which injected English context into
non-English conversations and muddied the continuation experience.
Adds a one-sentence instruction to the shared `_summarizer_preamble`
used by both the initial-compaction and iterative-update prompt paths.
Placing it in the preamble (rather than adding it separately to each
prompt) means both code paths stay in sync with one edit.
Ported from anomalyco/opencode#20581. The original PR (#4670) landed
before main's prompt templates were refactored to share the
`_summarizer_preamble` and `_template_sections` blocks, so the
cherry-pick conflicted on the now-obsolete inline sections; re-applied
the essential one-line change on top of the current structure.
Verified: 48/48 existing compressor tests pass.
Codex OAuth refresh tokens are single-use and rotate on every refresh.
Sharing them with the Codex CLI / VS Code via ~/.codex/auth.json made
concurrent use of both tools a race: whoever refreshed last invalidated
the other side's refresh_token. On top of that, the silent auto-import
path picked up placeholder / aborted-auth data from ~/.codex/auth.json
(e.g. literal {"access_token":"access-new","refresh_token":"refresh-new"})
and seeded it into the Hermes pool as an entry the selector could
eventually pick.
Hermes now owns its own Codex auth state end-to-end:
Removed
- agent/credential_pool.py: _sync_codex_entry_from_cli() method,
its pre-refresh + retry + _available_entries call sites, and the
post-refresh write-back to ~/.codex/auth.json.
- agent/credential_pool.py: auto-import from ~/.codex/auth.json in
_seed_from_singletons() — users now run `hermes auth openai-codex`
explicitly.
- hermes_cli/auth.py: silent runtime migration in
resolve_codex_runtime_credentials() — now surfaces
`codex_auth_missing` directly (message already points to `hermes auth`).
- hermes_cli/auth.py: post-refresh write-back in
_refresh_codex_auth_tokens().
- hermes_cli/auth.py: dead helper _write_codex_cli_tokens() and its 4
tests in test_auth_codex_provider.py.
Kept
- hermes_cli/auth.py: _import_codex_cli_tokens() — still used by the
interactive `hermes auth openai-codex` setup flow for a user-gated
one-time import (with "a separate login is recommended" messaging).
User-visible impact
- On existing installs with Hermes auth already present: no change.
- On a fresh install where the user has only logged in via Codex CLI:
`hermes chat --provider openai-codex` now fails with "No Codex
credentials stored. Run `hermes auth` to authenticate." The
interactive setup flow then detects ~/.codex/auth.json and offers a
one-time import.
- On an install where Codex CLI later refreshes its token: Hermes is
unaffected (we no longer read from that file at runtime).
Tests
- tests/hermes_cli/test_auth_codex_provider.py: 15/15 pass.
- tests/hermes_cli/test_auth_commands.py: 20/20 pass.
- tests/agent/test_credential_pool.py: 31/31 pass.
- Live E2E on openai-codex/gpt-5.4: 1 API call, 1.7s latency,
3 log lines, no refresh events, no auth drama.
The related 14:52 refresh-loop bug (hundreds of rotations/minute on a
single entry) is a separate issue — that requires a refresh-attempt
cap on the auth-recovery path in run_agent.py, which remains open.
Pass 3 of `_prune_old_tool_results` previously shrunk long `function.arguments`
blobs by slicing the raw JSON string at byte 200 and appending the literal
text `...[truncated]`. That routinely produced payloads like::
{"path": "/foo.md", "content": "# Long markdown
...[truncated]
— an unterminated string with no closing brace. Strict providers (observed
on MiniMax) reject this as `invalid function arguments json string` with a
non-retryable 400. Because the broken call survives in the session history,
every subsequent turn re-sends the same malformed payload and gets the same
400, locking the session into a re-send loop until the call falls out of
the window.
Fix: parse the arguments first, shrink long string leaves inside the parsed
structure, and re-serialise. Non-string values (paths, ints, booleans, lists)
pass through intact. Arguments that are not valid JSON to begin with (rare,
some backends use non-JSON tool args) are returned unchanged rather than
replaced with something neither we nor the provider can parse.
Observed in the wild: a `write_file` with ~800 chars of markdown `content`
triggered this on a real session against MiniMax-M2.7; every turn after
compression got rejected until the session was manually reset.
Tests:
- 7 direct tests of `_truncate_tool_call_args_json` covering valid-JSON
output, non-JSON pass-through, nested structures, non-string leaves,
scalar JSON, and Unicode preservation
- 1 end-to-end test through `_prune_old_tool_results` Pass 3 that
reproduces the exact failure payload shape from the incident
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
* fix(kimi): force fixed temperature on kimi-k2.* models (k2.5, thinking, turbo)
The prior override only matched the literal model name "kimi-for-coding",
but Moonshot's coding endpoint is hit with real model IDs such as
`kimi-k2.5`, `kimi-k2-turbo-preview`, `kimi-k2-thinking`, etc. Those
requests bypassed the override and kept the caller's temperature, so
Moonshot returns HTTP 400 "invalid temperature: only 0.6 is allowed for
this model" (or 1.0 for thinking variants).
Match the whole kimi-k2.* family:
* kimi-k2-thinking / kimi-k2-thinking-turbo -> 1.0 (thinking mode)
* all other kimi-k2.* -> 0.6 (non-thinking / instant mode)
Also accept an optional vendor prefix (e.g. `moonshotai/kimi-k2.5`) so
aggregator routings are covered.
* refactor(kimi): whitelist-match kimi coding models instead of prefix
Addresses review feedback on PR #12144.
- Replace `startswith("kimi-k2")` with explicit frozensets sourced from
Moonshot's kimi-for-coding model list. The prefix match would have also
clamped `kimi-k2-instruct` / `kimi-k2-instruct-0905`, which are the
separate non-coding K2 family with variable temperature (recommended 0.6
but not enforced — see huggingface.co/moonshotai/Kimi-K2-Instruct).
- Confirmed via platform.kimi.ai docs that all five coding models
(k2.5, k2-turbo-preview, k2-0905-preview, k2-thinking, k2-thinking-turbo)
share the fixed-temperature lock, so the preview-model mapping is no
longer an assumption.
- Drop the fragile `"thinking" in bare` substring test for a set lookup.
- Log a debug line on each override so operators can see when Hermes
silently rewrites temperature.
- Update class docstring. Extend the negative test to parametrize over
kimi-k2-instruct, Kimi-K2-Instruct-0905, and a hypothetical future
kimi-k2-experimental name — all must keep the caller's temperature.
persist_nous_credentials() now accepts an optional label kwarg which
gets embedded in providers.nous under the 'label' key.
_seed_from_singletons() prefers the embedded label over the
auto-derived label_from_token() fingerprint when materialising the
pool entry, so re-seeding on every load_pool('nous') preserves the
user's chosen label.
auth_commands.py threads --label through to the helper, restoring
parity with how other OAuth providers (anthropic, codex, google,
qwen) honor the flag.
Tests: 4 new (embed, reseed-survives, no-label fallback, end-to-end
through auth_add_command). All 390 nous/auth/credential_pool tests
pass.
Before: aggregator users (OpenRouter / Nous Portal) running 'auto'
routing for auxiliary tasks — compression, vision, web extraction,
session search, etc. — got routed to a cheap provider-side default
model (Gemini Flash). Non-aggregator users already got their main
model. Behavior was inconsistent and surprising — users picked
Claude / GPT / their preferred model, but side tasks ran on
Gemini Flash.
After: 'auto' means "use my main chat model" for every user,
regardless of provider type. Only when the main provider has no
working client does the fallback chain run (OpenRouter → Nous →
custom → Codex → API-key providers). Explicit per-task overrides
in config.yaml (auxiliary.<task>.provider / .model) still win —
they are a hard constraint, not subject to the auto policy.
Vision auto-detection follows the same policy: try main provider +
main model first (with _PROVIDER_VISION_MODELS overrides preserved
for providers like xiaomi and zai that ship a dedicated multimodal
model distinct from their chat model). Aggregator strict vision
backends are fallbacks, not the primary path.
Changes:
- agent/auxiliary_client.py: _resolve_auto() drops the
`_AGGREGATOR_PROVIDERS` guard. resolve_vision_provider_client()
auto branch unifies aggregator and exotic-provider paths —
everyone goes through resolve_provider_client() with main_model.
Dead _AGGREGATOR_PROVIDERS constant removed (was only used by
the guard we just removed).
- hermes_cli/main.py: aux config menu copy updated to reflect
the new semantics ("'auto' means 'use my main model'").
- tests/agent/test_auxiliary_main_first.py: 12 regression tests
covering OpenRouter/Nous/DeepSeek main paths, runtime-override
wins, explicit-config wins, vision override preservation for
exotic providers, and fallback-chain activation when the main
provider has no working client.
Co-authored-by: teknium1 <teknium@nousresearch.com>
build_skills_system_prompt() was using the skill directory name (skill_name)
when appending to skills_by_category in all three code paths (snapshot cache,
cold filesystem scan, external dirs). This meant any skill whose directory name
differed from its frontmatter `name` field would appear under the wrong name in
the system prompt, causing LLM routing failures.
The snapshot entry already stores both skill_name (dir) and frontmatter_name
(declared); switch the three tuple appends to use frontmatter_name. Also fix
the external-dir dedup set (seen_skill_names) to track frontmatter names for
consistency with the local-skill tuples now stored under frontmatter_name.
Fixes#11777
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Google-side 429 Code Assist errors now flow through Hermes' normal rate-limit
path (status_code on the exception, Retry-After preserved via error.response)
instead of being opaque RuntimeErrors. User sees a one-line capacity message
instead of a 500-char JSON dump.
Changes
- CodeAssistError grows status_code / response / retry_after / details attrs.
_extract_status_code in error_classifier picks up status_code and classifies
429 as FailoverReason.rate_limit, so fallback_providers triggers the same
way it does for SDK errors. run_agent.py line ~10428 already walks
error.response.headers for Retry-After — preserving the response means that
path just works.
- _gemini_http_error parses the Google error envelope (error.status +
error.details[].reason from google.rpc.ErrorInfo, retryDelay from
google.rpc.RetryInfo). MODEL_CAPACITY_EXHAUSTED / RESOURCE_EXHAUSTED / 404
model-not-found each produce a human-readable message; unknown shapes fall
back to the previous raw-body format.
- Drop gemma-4-26b-it from hermes_cli/models.py, hermes_cli/setup.py, and
agent/model_metadata.py — Google returned 404 for it today in local repro.
Kept gemma-4-31b-it (capacity-constrained but not retired).
Validation
| | Before | After |
|---------------------------|--------------------------------|-------------------------------------------|
| Error message | 'Code Assist returned HTTP 429: {500 chars JSON}' | 'Gemini capacity exhausted for gemini-2.5-pro (Google-side throttle...)' |
| status_code on error | None (opaque RuntimeError) | 429 |
| Classifier reason | unknown (string-match fallback) | FailoverReason.rate_limit |
| Retry-After honored | ignored | extracted from RetryInfo or header |
| gemma-4-26b-it picker | advertised (404s on Google) | removed |
Unit + E2E tests cover non-streaming 429, streaming 429, 404 model-not-found,
Retry-After header fallback, malformed body, and classifier integration.
Targeted suites: tests/agent/test_gemini_cloudcode.py (81 tests), full
tests/hermes_cli (2203 tests) green.
Co-authored-by: teknium1 <teknium@nousresearch.com>
Follow-up on the native NVIDIA NIM provider salvage. The original PR wired
PROVIDER_REGISTRY + HERMES_OVERLAYS correctly but missed several touchpoints
required for full parity with other OpenAI-compatible providers (xai,
huggingface, deepseek, zai).
Gaps closed:
- hermes_cli/main.py:
- Add 'nvidia' to the _model_flow_api_key_provider dispatch tuple so
selecting 'NVIDIA NIM' in `hermes model` actually runs the api-key
provider flow (previously fell through silently).
- Add 'nvidia' to `hermes chat --provider` argparse choices so the
documented test command (`hermes chat --provider nvidia --model ...`)
parses successfully.
- hermes_cli/config.py: Register NVIDIA_API_KEY and NVIDIA_BASE_URL in
OPTIONAL_ENV_VARS so setup wizard can prompt for them and they're
auto-added to the subprocess env blocklist.
- hermes_cli/doctor.py: Add NVIDIA NIM row to `_apikey_providers` so
`hermes doctor` probes https://integrate.api.nvidia.com/v1/models.
- hermes_cli/dump.py: Add NVIDIA_API_KEY → 'nvidia' mapping for
`hermes dump` credential masking.
- tests/tools/test_local_env_blocklist.py: Extend registry_vars fixture
with NVIDIA_API_KEY to verify it's blocked from leaking into subprocesses.
- agent/model_metadata.py: Add 'nemotron' → 131072 context-length entry
so all Nemotron variants get 128K context via substring match (rather
than falling back to MINIMUM_CONTEXT_LENGTH).
- hermes_cli/models.py: Fix hallucinated model ID
'nvidia/nemotron-3-nano-8b-a4b' → 'nvidia/nemotron-3-nano-30b-a3b'
(verified against live integrate.api.nvidia.com/v1/models catalog).
Expand curated list from 5 to 9 agentic models mapping to OpenRouter
defaults per provider-guide convention: add qwen3.5-397b-a17b,
deepseek-v3.2, llama-3.3-nemotron-super-49b-v1.5, gpt-oss-120b.
- cli-config.yaml.example: Document 'nvidia' provider option.
- scripts/release.py: Map asurla@nvidia.com → anniesurla in AUTHOR_MAP
for CI attribution.
E2E verified: `hermes chat --provider nvidia ...` now reaches NVIDIA's
endpoint (returns 401 with bogus key instead of argparse error);
`hermes doctor` detects NVIDIA NIM when NVIDIA_API_KEY is set.
Adds NVIDIA NIM as a first-class provider: ProviderConfig in
auth.py, HermesOverlay in providers.py, curated models
(Nemotron plus other open source models hosted on
build.nvidia.com), URL mapping in model_metadata.py, aliases
(nim, nvidia-nim, build-nvidia, nemotron), and env var tests.
Docs updated: providers page, quickstart table, fallback
providers table, and README provider list.
* feat(skills): add 'hermes skills reset' to un-stick bundled skills
When a user edits a bundled skill, sync flags it as user_modified and
skips it forever. The problem: if the user later tries to undo the edit
by copying the current bundled version back into ~/.hermes/skills/, the
manifest still holds the old origin hash from the last successful
sync, so the fresh bundled hash still doesn't match and the skill stays
stuck as user_modified.
Adds an escape hatch for this case.
hermes skills reset <name>
Drops the skill's entry from ~/.hermes/skills/.bundled_manifest and
re-baselines against the user's current copy. Future 'hermes update'
runs accept upstream changes again. Non-destructive.
hermes skills reset <name> --restore
Also deletes the user's copy and re-copies the bundled version.
Use when you want the pristine upstream skill back.
Also available as /skills reset in chat.
- tools/skills_sync.py: new reset_bundled_skill(name, restore=False)
- hermes_cli/skills_hub.py: do_reset() + wired into skills_command and
handle_skills_slash; added to the slash /skills help panel
- hermes_cli/main.py: argparse entry for 'hermes skills reset'
- tests/tools/test_skills_sync.py: 5 new tests covering the stuck-flag
repro, --restore, unknown-skill error, upstream-removed-skill, and
no-op on already-clean state
- website/docs/user-guide/features/skills.md: new 'Bundled skill updates'
section explaining the origin-hash mechanic + reset usage
* fix(auth): codex auth remove no longer silently undone by auto-import
'hermes auth remove openai-codex' appeared to succeed but the credential
reappeared on the next command. Two compounding bugs:
1. _seed_from_singletons() for openai-codex unconditionally re-imports
tokens from ~/.codex/auth.json whenever the Hermes auth store is
empty (by design — the Codex CLI and Hermes share that file). There
was no suppression check, unlike the claude_code seed path.
2. auth_remove_command's cleanup branch only matched
removed.source == 'device_code' exactly. Entries added via
'hermes auth add openai-codex' have source 'manual:device_code', so
for those the Hermes auth store's providers['openai-codex'] state was
never cleared on remove — the next load_pool() re-seeded straight
from there.
Net effect: there was no way to make a codex removal stick short of
manually editing both ~/.hermes/auth.json and ~/.codex/auth.json before
opening Hermes again.
Fix:
- Add unsuppress_credential_source() helper (mirrors
suppress_credential_source()).
- Gate the openai-codex branch in _seed_from_singletons() with
is_source_suppressed(), matching the claude_code pattern.
- Broaden auth_remove_command's codex match to handle both
'device_code' and 'manual:device_code' (via endswith check), always
call suppress_credential_source(), and print guidance about the
unchanged ~/.codex/auth.json file.
- Clear the suppression marker in auth_add_command's openai-codex
branch so re-linking via 'hermes auth add openai-codex' works.
~/.codex/auth.json is left untouched — that's the Codex CLI's own
credential store, not ours to delete.
Tests cover: unsuppress helper behavior, remove of both source
variants, add clears suppression, seed respects suppression. E2E
verified: remove → load → add → load flow now behaves correctly.
The cache-read, cache-write, and total estimated-cost values shown in
/insights (and the per-model Cost column) were unreliable. Hide them from
both terminal and gateway renderings.
The underlying data pipeline is untouched — sessions still store
cache_read_tokens, cache_write_tokens, and estimated_cost_usd; the web
server, /usage command, and status bar are unaffected. Only the
InsightsEngine display layer is trimmed.
Changes:
- format_terminal: drop 'Cache read / Cache write' line, drop 'Est. cost'
from the Total tokens row, drop per-model 'Cost' column, drop the
'* Cost N/A for custom/self-hosted' footnote.
- format_gateway: drop cache breakdown from Tokens line, drop 'Est. cost'
line, drop per-model cost suffix.
- Tests updated to assert these strings are now absent.
run_agent.py passes httpx.Timeout(connect=30, read=120, write=1800,
pool=30) as the timeout kwarg on the streaming path. The OpenAI SDK
handles this natively, but CopilotACPClient._create_chat_completion()
called float(timeout or default), which raises TypeError because
httpx.Timeout doesn't implement __float__.
Normalize the timeout before passing to _run_prompt: plain floats/ints
pass through, httpx.Timeout objects get their largest component
extracted (write=1800s is the correct wall-clock budget for the ACP
subprocess), and None falls back to the 900s default.
Regression from #11161 (Claude Opus 4.7 migration, commit 0517ac3e).
The Opus 4.7 migration changed `ADAPTIVE_EFFORT_MAP["xhigh"]` from "max"
(the pre-migration alias) to "xhigh" to preserve the new 4.7 effort level
as distinct from max. This is correct for 4.7, but Opus/Sonnet 4.6 only
expose 4 levels (low/medium/high/max) — sending "xhigh" there now 400s:
BadRequestError [HTTP 400]: This model does not support effort
level 'xhigh'. Supported levels: high, low, max, medium.
Users who set reasoning_effort=xhigh as their default (xhigh is the
recommended default for coding/agentic on 4.7 per the Anthropic migration
guide) now 400 every request the moment they switch back to a 4.6 model
via `/model` or config. Verified live against the Anthropic API on
`anthropic==0.94.0`.
Fix: make the mapping model-aware. Add `_supports_xhigh_effort()`
predicate (matches 4-7/4.7 substrings, mirroring the existing
`_supports_adaptive_thinking` / `_forbids_sampling_params` pattern).
On pre-4.7 adaptive models, downgrade xhigh→max (the strongest effort
those models accept, restoring pre-migration behavior). On 4.7+, keep
xhigh as a distinct level.
Per Anthropic's migration guide, xhigh is 4.7-only:
https://platform.claude.com/docs/en/about-claude/models/migration-guide
> Opus 4.7 effort levels: max, xhigh (new), high, medium, low.
> Opus 4.6 effort levels: max, high, medium, low.
SDK typing confirms: `anthropic.types.OutputConfigParam.effort: Literal[
"low", "medium", "high", "max"]` (v0.94.0 not yet updated for xhigh).
## Test plan
Verified live on macOS 15.5 / anthropic==0.94.0:
claude-opus-4-6 + effort=xhigh → output_config.effort=max → 200 OK
claude-opus-4-7 + effort=xhigh → output_config.effort=xhigh → 200 OK
claude-opus-4-6 + effort=max → output_config.effort=max → 200 OK
claude-opus-4-7 + effort=max → output_config.effort=max → 200 OK
`tests/agent/test_anthropic_adapter.py` — 120 pass (replaced 1 bugged
test that asserted the broken behavior, added 1 for 4.7 preservation).
Full adapter suite: 120 passed in 1.05s.
Broader suite (agent + run_agent + cli/gateway reasoning): 2140 passed
(2 pre-existing failures on clean upstream/main, unrelated).
## Platforms
Tested on macOS 15.5. No platform-specific code paths touched.
Claude Opus 4.7 introduced several breaking API changes that the current
codebase partially handled but not completely. This patch finishes the
migration per the official migration guide at
https://platform.claude.com/docs/en/about-claude/models/migration-guideFixesNousResearch/hermes-agent#11137
Breaking-change coverage:
1. Adaptive thinking + output_config.effort — 4.7 is now recognized by
_supports_adaptive_thinking() (extends previous 4.6-only gate).
2. Sampling parameter stripping — 4.7 returns 400 for any non-default
temperature / top_p / top_k. build_anthropic_kwargs drops them as a
safety net; the OpenAI-protocol auxiliary path (_build_call_kwargs)
and AnthropicCompletionsAdapter.create() both early-exit before
setting temperature for 4.7+ models. This keeps flush_memories and
structured-JSON aux paths that hardcode temperature from 400ing
when the aux model is flipped to 4.7.
3. thinking.display = "summarized" — 4.7 defaults display to "omitted",
which silently hides reasoning text from Hermes's CLI activity feed
during long tool runs. Restoring "summarized" preserves 4.6 UX.
4. Effort level mapping — xhigh now maps to xhigh (was xhigh→max, which
silently over-efforted every coding/agentic request). max is now a
distinct ceiling per Anthropic's 5-level effort model.
5. New stop_reason values — refusal and model_context_window_exceeded
were silently collapsed to "stop" (end_turn) by the adapter's
stop_reason_map. Now mapped to "content_filter" and "length"
respectively, matching upstream finish-reason handling already in
bedrock_adapter.
6. Model catalogs — claude-opus-4-7 added to the Anthropic provider
list, anthropic/claude-opus-4.7 added at top of OpenRouter fallback
catalog (recommended), claude-opus-4-7 added to model_metadata
DEFAULT_CONTEXT_LENGTHS (1M, matching 4.6 per migration guide).
7. Prefill docstrings — run_agent.AIAgent and BatchRunner now document
that Anthropic Sonnet/Opus 4.6+ reject a trailing assistant-role
prefill (400).
8. Tests — 4 new tests in test_anthropic_adapter covering display
default, xhigh preservation, max on 4.7, refusal / context-overflow
stop_reason mapping, plus the sampling-param predicate. test_model_metadata
accepts 4.7 at 1M context.
Tested on macOS 15.5 (darwin). 119 tests pass in
tests/agent/test_anthropic_adapter.py, 1320 pass in tests/agent/.
Ensure _align_boundary_backward never pushes the last user message
into the compressed region. Without this, compression could delete
the user active task instruction mid-session.
Cherry-picked from #10969 by @sontianye. Fixes#10896.
resolve_vision_provider_client() was receiving the raw call_llm
parameters instead of the resolved provider/model/key/url from
_resolve_task_provider_model(). This caused config overrides
(auxiliary.vision.provider, etc.) to be silently discarded.
Cherry-picked from #10901 by @lrawnsley.
The gateway compression notifications were already removed in commit cc63b2d1
(PR #4139), but the agent-level context pressure warnings (85%/95% tiered
alerts via _emit_context_pressure) were still firing on both CLI and gateway.
Removed:
- _emit_context_pressure method and all call sites in run_conversation()
- Class-level dedup state (_context_pressure_last_warned, _CONTEXT_PRESSURE_COOLDOWN)
- Instance attribute _context_pressure_warned_at
- Pressure reset logic in _compress_context
- format_context_pressure and format_context_pressure_gateway from agent/display.py
- Orphaned ANSI constants that only served these functions
- tests/run_agent/test_context_pressure.py (all 361 lines)
Compression itself continues to run silently in the background.
Closes#3784
Skins define waiting_faces, thinking_faces, and thinking_verbs in their
spinner config, but all 7 call sites in run_agent.py used hardcoded class
constants. Add three classmethods on KawaiiSpinner that query the active
skin first and fall back to the class constants, matching the existing
pattern used for wings/tool_prefix/tool_emojis.
Co-authored-by: nosleepcassette <nosleepcassette@users.noreply.github.com>
_load_skill_payload() reconstructed skill_dir as SKILLS_DIR / relative_path,
which is wrong for external skills from skills.external_dirs — they live
outside SKILLS_DIR entirely. Scripts and linked files failed to load.
Fix: skill_view() now includes the absolute skill_dir in its result dict.
_load_skill_payload() uses that directly when available, falling back to
the SKILLS_DIR-relative reconstruction only for legacy responses.
Closes#10313
When Nous returns a 429, the retry amplification chain burns up to 9
API requests per conversation turn (3 SDK retries × 3 Hermes retries),
each counting against RPH and deepening the rate limit. With multiple
concurrent sessions (cron + gateway + auxiliary), this creates a spiral
where retries keep the limit tapped indefinitely.
New module: agent/nous_rate_guard.py
- Shared file-based rate limit state (~/.hermes/rate_limits/nous.json)
- Parses reset time from x-ratelimit-reset-requests-1h, x-ratelimit-
reset-requests, retry-after headers, or error context
- Falls back to 5-minute default cooldown if no header data
- Atomic writes (tempfile + rename) for cross-process safety
- Auto-cleanup of expired state files
run_agent.py changes:
- Top-of-retry-loop guard: when another session already recorded Nous
as rate-limited, skip the API call entirely. Try fallback provider
first, then return a clear message with the reset time.
- On 429 from Nous: record rate limit state and skip further retries
(sets retry_count = max_retries to trigger fallback path)
- On success from Nous: clear the rate limit state so other sessions
know they can resume
auxiliary_client.py changes:
- _try_nous() checks rate guard before attempting Nous in the auxiliary
fallback chain. When rate-limited, returns (None, None) so the chain
skips to the next provider instead of piling more requests onto Nous.
This eliminates three sources of amplification:
1. Hermes-level retries (saves 6 of 9 calls per turn)
2. Cross-session retries (cron + gateway all skip Nous)
3. Auxiliary fallback to Nous (compression/session_search skip too)
Includes 24 tests covering the rate guard module, header parsing,
state lifecycle, and auxiliary client integration.
When proxy env vars (HTTP_PROXY, HTTPS_PROXY, ALL_PROXY) contain
malformed URLs — e.g. 'http://127.0.0.1:6153export' from a broken
shell config — the OpenAI/httpx client throws a cryptic 'Invalid port'
error that doesn't identify the offending variable.
Add _validate_proxy_env_urls() and _validate_base_url() in
auxiliary_client.py, called from resolve_provider_client() and
_create_openai_client() to fail fast with a clear, actionable error
message naming the broken env var or URL.
Closes#6360
Co-authored-by: MestreY0d4-Uninter <MestreY0d4-Uninter@users.noreply.github.com>
Found via trace data audit: JWT tokens (eyJ...) and Discord snowflake
mentions (<@ID>) were passing through unredacted.
JWT pattern: matches 1/2/3-part tokens starting with eyJ (base64 for '{').
Zero false-positive risk — no normal text matches eyJ + 10+ base64url chars.
Discord pattern: matches <@digits> and <@!digits> with 17-20 digit snowflake
IDs. Syntactically unique to Discord's mention format.
Both patterns follow the same structural-uniqueness standard as existing
prefix patterns (sk-, ghp_, AKIA, etc.).
