feat(i18n): add display.language for static message translation (zh/ja/de/es) by teknium1 · Pull Request #20231 · NousResearch/hermes-agent

teknium1 · 2026-05-05T13:04:35Z

Summary

Adds a thin i18n layer so the CLI approval prompt and a handful of gateway replies render in Chinese, Japanese, German, or Spanish when display.language is set. Agent-generated output, logs, tool results, and slash-command descriptions stay English on purpose — see scope below.

Scope (intentional thin slice)

Translated (~19 strings per locale):

CLI dangerous-command approval prompt — header, choice menu, prompt text, all outcome lines (allowed once/session/always, denied, cancelled, timeout)
Gateway replies: restart-drain notice, goal cleared / no active goal, approval expired, config read/save errors

NOT translated (stays English):

Agent responses (controlled by the model + user prompt, not this layer)
Log lines, error tracebacks, debug output
Slash-command canonical names and descriptions
Tool outputs
Setup wizard, /help text, banner, spinner verbs

If a user wants the agent itself to reply in another language, they say so in their prompt — orthogonal to this.

Why thin slice

Full i18n across cli.py / gateway/run.py / hermes_cli/setup.py would mean extracting hundreds of English literals and paying a forever-tax to keep new PRs wrapping strings in t(...). Scope was reviewed before building — thin slice covers the highest-impact visible messages without the drift tax.

Changes

agent/i18n.py: catalog loader + t() helper + language resolution (HERMES_LANGUAGE env > config > en default). Process-cached. Supports aliases (chinese → zh, zh-CN → zh, etc.)
locales/{en,zh,ja,de,es}.yaml: source catalog (en) + 4 translations
hermes_cli/config.py: display.language default entry
tools/approval.py: CLI approval prompt wired through t()
gateway/run.py: 6 high-impact reply sites wired through t()
website/docs/user-guide/configuration.md: documents the setting plus its explicit scope
tests/agent/test_i18n.py: 21 tests

Validation

	Result
`tests/agent/test_i18n.py`	21/21 passed
`tests/tools/test_approval.py`	132/132 passed
`tests/hermes_cli/test_config.py`	52/52 passed
E2E: `display.language: zh` → live `t()`	Chinese output verified for all 7 wired messages
E2E: `HERMES_LANGUAGE=ja` override	Japanese, env beats config
E2E: no config + no env	English (default fallback)

Key design choices

Placeholder parity test — catches typos like {descricao} for {description} at test time, not runtime
Key parity test — adding a key to en.yaml without adding it to every other locale fails CI, preventing silent English-only regression
Graceful fallback — missing key in locale → English; missing in English too → the key path itself (never crashes)
No new env var for config — per the .env is for secrets only rule, HERMES_LANGUAGE is an internal/test override; display.language in config.yaml is the user-facing knob
Process-cached config read — t() is called in hot paths; clearable via reset_language_cache() if we ever need runtime re-read

Future extension

The catalog format supports nesting and arbitrary key paths, so follow-up PRs can expand coverage (setup wizard, /help, setup summary) by adding keys to all 5 YAML files in one change — no core-code churn per language. Community translators can contribute new languages by adding one more YAML file.

Test plan

./scripts/run_tests.sh tests/agent/test_i18n.py tests/tools/test_approval.py tests/hermes_cli/test_config.py
Set display.language: zh in config.yaml → trigger a dangerous command in CLI → prompt shows Chinese
Set HERMES_LANGUAGE=ja → same, Japanese
Unset both → English baseline

Removes the _detect_stale_code / _trigger_stale_code_restart mechanism introduced in #17648 and iterated in #19740. On every incoming message the gateway compared the boot-time git HEAD SHA to the current SHA on disk, and if they differed it would reply with Gateway code was updated in the background -- restarting this gateway so your next message runs on the new code. Please retry in a moment. and then kick off a graceful restart. This is unwanted behaviour: users who run a long-lived gateway and do their own ad-hoc git operations on the checkout end up with their chat interrupted and the current message dropped every time HEAD moves, with no way to opt out. If an operator really needs the old protection against stale sys.modules after "hermes update", the SIGKILL-survivor sweep in hermes update (hermes_cli/main.py, also tagged #17648) already handles the supervisor-respawn case on its own. Removed: gateway/run.py: - _STALE_CODE_SENTINELS, _GIT_SHA_CACHE_TTL_SECS - _read_git_head_sha(), _compute_repo_mtime() module helpers - class-level _boot_wall_time / _boot_repo_mtime / _boot_git_sha / _stale_code_restart_triggered defaults - __init__ boot-snapshot block (_boot_*, _cached_current_sha*, _repo_root_for_staleness, _stale_code_notified) - _current_git_sha_cached(), _detect_stale_code(), _trigger_stale_code_restart() methods - stale-code check + user-facing restart notice at the top of _handle_message() tests/gateway/test_stale_code_self_check.py (deleted, 412 lines) No new logic added. Zero remaining references to any removed symbol. Gateway test suite passes the same 4589 tests it passed before; the 3 pre-existing unrelated failures (discord free-channel, feishu bot admission, teams typing) are unchanged by this commit.

…a/de/es) Adds a thin-slice i18n layer covering the highest-impact static user-facing messages: the CLI dangerous-command approval prompt and a handful of gateway slash-command replies (restart-drain, goal cleared, approval expired, config read/save errors). Out of scope (stays English): agent responses, log lines, tool outputs, slash-command descriptions, error tracebacks. Infrastructure: - agent/i18n.py: catalog loader, t() helper, language resolution (HERMES_LANGUAGE env var > display.language config > en) - locales/{en,zh,ja,de,es}.yaml: ~19 translated strings per language - display.language in DEFAULT_CONFIG (hermes_cli/config.py) Tests: - tests/agent/test_i18n.py: 21 tests covering catalog parity, placeholder parity across locales, fallback behavior, env-var override, alias normalization, missing-key graceful degradation. Docs: - website/docs/user-guide/configuration.md: display.language entry plus a short section explaining scope so users don't expect agent responses to translate via this knob.

LeMoussel · 2026-05-05T15:46:13Z

No French translation?
I'm a French user of hermes-agent. I can help you add a French translation.

Foolafroos · 2026-05-05T17:29:35Z

No French translation? I'm a French user of hermes-agent. I can help you add a French translation.

#20329 Ici mon ami <3

ebourgess · 2026-05-06T07:59:50Z

No arabic though? tsk tsk

S17S17 · 2026-05-06T12:06:41Z

Noo arabic ?

woojinstartup · 2026-05-07T02:03:41Z

lfg Korean

…a/de/es) (NousResearch#20231) * revert(gateway): remove stale-code self-check and auto-restart Removes the _detect_stale_code / _trigger_stale_code_restart mechanism introduced in NousResearch#17648 and iterated in NousResearch#19740. On every incoming message the gateway compared the boot-time git HEAD SHA to the current SHA on disk, and if they differed it would reply with Gateway code was updated in the background -- restarting this gateway so your next message runs on the new code. Please retry in a moment. and then kick off a graceful restart. This is unwanted behaviour: users who run a long-lived gateway and do their own ad-hoc git operations on the checkout end up with their chat interrupted and the current message dropped every time HEAD moves, with no way to opt out. If an operator really needs the old protection against stale sys.modules after "hermes update", the SIGKILL-survivor sweep in hermes update (hermes_cli/main.py, also tagged NousResearch#17648) already handles the supervisor-respawn case on its own. Removed: gateway/run.py: - _STALE_CODE_SENTINELS, _GIT_SHA_CACHE_TTL_SECS - _read_git_head_sha(), _compute_repo_mtime() module helpers - class-level _boot_wall_time / _boot_repo_mtime / _boot_git_sha / _stale_code_restart_triggered defaults - __init__ boot-snapshot block (_boot_*, _cached_current_sha*, _repo_root_for_staleness, _stale_code_notified) - _current_git_sha_cached(), _detect_stale_code(), _trigger_stale_code_restart() methods - stale-code check + user-facing restart notice at the top of _handle_message() tests/gateway/test_stale_code_self_check.py (deleted, 412 lines) No new logic added. Zero remaining references to any removed symbol. Gateway test suite passes the same 4589 tests it passed before; the 3 pre-existing unrelated failures (discord free-channel, feishu bot admission, teams typing) are unchanged by this commit. * feat(i18n): add display.language for static message translation (zh/ja/de/es) Adds a thin-slice i18n layer covering the highest-impact static user-facing messages: the CLI dangerous-command approval prompt and a handful of gateway slash-command replies (restart-drain, goal cleared, approval expired, config read/save errors). Out of scope (stays English): agent responses, log lines, tool outputs, slash-command descriptions, error tracebacks. Infrastructure: - agent/i18n.py: catalog loader, t() helper, language resolution (HERMES_LANGUAGE env var > display.language config > en) - locales/{en,zh,ja,de,es}.yaml: ~19 translated strings per language - display.language in DEFAULT_CONFIG (hermes_cli/config.py) Tests: - tests/agent/test_i18n.py: 21 tests covering catalog parity, placeholder parity across locales, fallback behavior, env-var override, alias normalization, missing-key graceful degradation. Docs: - website/docs/user-guide/configuration.md: display.language entry plus a short section explaining scope so users don't expect agent responses to translate via this knob.

teknium1 added 2 commits May 5, 2026 03:44

alt-glitch added type/feature New feature or request P3 Low — cosmetic, nice to have comp/cli CLI entry point, hermes_cli/, setup wizard comp/gateway Gateway runner, session dispatch, delivery area/config Config system, migrations, profiles labels May 5, 2026

teknium1 merged commit 7de3c86 into main May 5, 2026
10 of 12 checks passed

teknium1 deleted the hermes/hermes-a8b6587e branch May 5, 2026 15:03

leprincep35700 mentioned this pull request May 5, 2026

feat(i18n): add Italian locale #20371

Closed

3 tasks

pedroborges mentioned this pull request May 5, 2026

feat(i18n): add Portuguese locale support #20443

Closed

github-actions Bot mentioned this pull request May 7, 2026

🦞 OpenClaw 生态日报 2026-05-07 ivanweng2077/big_model_radar#6

Open

BrewTestBot mentioned this pull request May 7, 2026

hermes-agent 2026.5.7 Homebrew/homebrew-core#281437

Merged

1 task

github-actions Bot mentioned this pull request May 8, 2026

chore: bump NousResearch/hermes-agent version from v2026.4.30 to v2026.5.7 Docker-Hub-sirmark/docker-hermes-agent#5

Merged

kshitijk4poor mentioned this pull request May 9, 2026

feat(i18n) Internationalize hardcoded status messages in run_agent.py #21639

Closed

subinium mentioned this pull request May 11, 2026

polish(core): expand SupportedLocale beyond en/ko — add zh, ja, de, es, fr, uk, tr (Hermes v0.13 parity) subinium/CrowClaw#335

Open

3 tasks

andreszb mentioned this pull request May 11, 2026

[Bug]: Nix package omits locales, causing raw i18n keys in gateway command replies #23943

Closed

XIYBHK mentioned this pull request May 12, 2026

Chinese (Simplified) localization — community patch set, offer to contribute #24252

Open

alt-glitch mentioned this pull request May 28, 2026

Feature Request: i18n / 多语言支持（国际化） #33749

Open

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat(i18n): add display.language for static message translation (zh/ja/de/es)#20231

feat(i18n): add display.language for static message translation (zh/ja/de/es)#20231
teknium1 merged 2 commits into
mainfrom
hermes/hermes-a8b6587e

teknium1 commented May 5, 2026

Uh oh!

Uh oh!

LeMoussel commented May 5, 2026

Uh oh!

Foolafroos commented May 5, 2026

Uh oh!

ebourgess commented May 6, 2026

Uh oh!

S17S17 commented May 6, 2026

Uh oh!

woojinstartup commented May 7, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

7 participants

Conversation

teknium1 commented May 5, 2026

Summary

Scope (intentional thin slice)

Why thin slice

Changes

Validation

Key design choices

Future extension

Test plan

Uh oh!

Uh oh!

LeMoussel commented May 5, 2026

Uh oh!

Foolafroos commented May 5, 2026

Uh oh!

ebourgess commented May 6, 2026

Uh oh!

S17S17 commented May 6, 2026

Uh oh!

woojinstartup commented May 7, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

7 participants