refactor(gateway): migrate Discord adapter to bundled plugin by kshitijk4poor · Pull Request #24356 · NousResearch/hermes-agent

kshitijk4poor · 2026-05-12T12:42:34Z

Summary

First migration of an existing built-in platform adapter to the bundled plugin system established by IRC / Teams / LINE / Google Chat. Closes #24325; advances the umbrella refactor in #3823.

This PR achieves complete Teams-style parity: the Discord adapter lives at plugins/platforms/discord/{__init__.py, adapter.py, plugin.yaml}, registers via register(ctx), and every hook Teams uses is wired (plus apply_yaml_config_fn from #25443, which Discord is the first consumer of). No back-compat shim at the old gateway/platforms/discord.py path — that file is deleted and all 78 consumer references are rewritten.

Why Discord first

All four existing platform plugins (irc, teams, line, google_chat) under plugins/platforms/ were born as plugins. Discord is the largest and most deeply-integrated built-in adapter — 5,101 LOC, 25 dedicated test files, ~1,976 Platform.DISCORD references in non-adapter code. If the migration pattern works for Discord, it works for everything else.

Hooks wired

Hook	Where the implementation lives	What it replaces in core
`check_fn=check_discord_requirements`	inside the adapter	n/a (already plugin-style)
`is_connected=_is_connected`	new helper	implicit check in `gateway/config.py`
`setup_fn=interactive_setup`	new helper (68 LOC moved from `hermes_cli/setup.py`)	`_setup_discord` + `_PLATFORMS["discord"]` dict
`standalone_sender_fn=_standalone_send`	new helper (190 LOC moved from `tools/send_message_tool.py`)	the `Platform.DISCORD` elif in `_send_to_platform`
`apply_yaml_config_fn=_apply_yaml_config` (NEW, #25443 consumer)	new helper (89 LOC)	59-LOC `discord_cfg` block in `gateway/config.py::load_gateway_config`
`allowed_users_env="DISCORD_ALLOWED_USERS"` + `allow_all_env="DISCORD_ALLOW_ALL_USERS"`	registry field	hardcoded entries in `_is_user_authorized` allowlist maps (still there — generic concern)
`cron_deliver_env_var="DISCORD_HOME_CHANNEL"`	registry field	hardcoded entry in `cron/scheduler.py` (still there — generic concern)
`max_message_length=2000`	registry field	`_MAX_LENGTHS` dict in `tools/send_message_tool.py`
`emoji="🎮"` + `required_env=["DISCORD_BOT_TOKEN"]` + `install_hint=...`	registry fields	various

What changed

File	Change
`gateway/platforms/discord.py` (5,101 LOC) → `plugins/platforms/discord/adapter.py`	git rename (R090 — large because we appended ~340 LOC of moved-in hook code)
`plugins/platforms/discord/{__init__.py, plugin.yaml}`	New plugin shell, declares `requires_env` / `optional_env`
`plugins/platforms/discord/adapter.py` (tail)	Append all hook implementations + `register(ctx)` block. The dataclass already supports every field needed; no registry API extension required (the `apply_yaml_config_fn` hook landed in #25443).
`gateway/config.py`	Delete the 59-LOC legacy `discord_cfg` YAML→env bridge block (moved into plugin's `_apply_yaml_config`)
`gateway/run.py`	Replace `Platform.DISCORD elif` branch (−9 LOC) with generic `gateway_runner` post-creation hook (+6 LOC)
`tools/send_message_tool.py`	Delete `_send_discord` + 4 helpers (221 LOC); replace `Platform.DISCORD` elif with 10-line registry-hook dispatch
`hermes_cli/setup.py`	Delete `_setup_discord` + `_clean_discord_user_ids` (68 LOC)
`hermes_cli/gateway.py`	Delete `_PLATFORMS["discord"]` dict (32 LOC) + remove `"discord": _s._setup_discord` from `_builtin_setup_fn`
28 test files + `tools/send_message_tool.py` + `hermes_cli/commands.py` docstring	Rewrite 78 references to old import path

Diff stats

38 files changed, +621 / −473 — net positive due to the YAML hook's 89 new LOC plus the standalone/setup hook bodies, but every line moved has a clear plugin home now. Git's rename detection hits at 90% similarity even after the new hooks were appended.

Test coverage

All 568 Discord-specific tests pass across 25 test_discord_*.py files plus voice/send/text-batching/reload-skills/stream-consumer/integration tests.
All 147 YAML-bridge-touching tests pass (test_discord_reply_mode, test_discord_free_response, test_discord_allowed_channels, test_discord_allowed_mentions, test_discord_channel_controls, test_discord_reactions, test_discord_thread_persistence, test_runtime_footer) — strongest signal that the YAML→env hook behaves identically to the legacy block.
Broader gateway/cron/integration sweep (1297 tests) introduces zero new failures vs main. The remaining pre-existing failures in tests/gateway/test_tts_media_routing.py and tests/e2e/test_platform_commands.py reproduce identically on the unchanged main revision (the latter affects all 3 platforms — telegram/discord/slack — for the same reason; not Discord-specific).
Plugin discovery sanity check confirms Discord registers alongside the other four platform plugins:

Registered platforms: ['discord', 'google_chat', 'irc', 'line', 'teams']

Out of scope (separate concerns, not Discord-specific)

The following Discord-shaped tendrils in core were deliberately not moved — they are generic platform-registry concerns affecting every platform:

gateway/config.py:1205 DISCORD_BOT_TOKEN → config.token env enablement — same shape Telegram has. The existing env_enablement_fn registry hook only seeds extra, not .token, so it can't replace this without an adapter refactor to read from extra["bot_token"].
gateway/run.py voice-mode hooks (self.adapters.get(Platform.DISCORD) for start_voice_mode/stop_voice_mode), role-based auth, DISCORD_ALLOW_BOTS branch in _is_user_authorized, _UPDATE_ALLOWED_PLATFORMS frozenset, and the per-platform allowlist maps — generic platform-registry concerns.
Platform.DISCORD enum literal — stable identifier used as dict keys throughout the codebase; removing it is a separate refactor with no real benefit.
tools/discord_tool.py and tools/environments/local.py — first-class agent tools and env-passthrough config, neither is the gateway adapter.

Each of these is worth its own scoping issue when the time comes — and is fair game once this lands.

cc @teknium1 — this concerns issue #3823 directly and is the first real consumer of the apply_yaml_config_fn hook from #25443.

Closes #24325
Refs #3823, #25443

Lets platform plugins own their YAML→env config bridge instead of forcing core gateway/config.py to know every platform's schema. The hook receives the full parsed config.yaml and the platform's own sub-dict, may mutate os.environ (env > YAML precedence preserved via the standard `not os.getenv(...)` guards), and may return a dict to merge into PlatformConfig.extra. It runs during load_gateway_config() after the existing generic shared-key loop and before _apply_env_overrides(), mirroring the env_enablement_fn dispatch pattern (NousResearch#21306, NousResearch#21331). Pure addition — no behavior change for existing platforms. Each of the eight platforms with hardcoded YAML→env blocks today (discord, telegram, whatsapp, slack, dingtalk, mattermost, matrix, feishu, ~252 LOC in gateway/config.py) can migrate in independent follow-up PRs; the hardcoded blocks remain functional in the meantime, and their `not os.getenv(...)` guards make them no-ops for any env var the hook already set. Test coverage: 10 new tests in tests/gateway/test_platform_registry.py covering field default, callable acceptance, env mutation, extras merge, both signature args, exception swallowing, missing/non-dict sections, and env > YAML precedence. Refs NousResearch#3823, NousResearch#24356. Closes NousResearch#24836.

Lets platform plugins own their YAML→env config bridge instead of forcing core gateway/config.py to know every platform's schema. The hook receives the full parsed config.yaml and the platform's own sub-dict, may mutate os.environ (env > YAML precedence preserved via the standard `not os.getenv(...)` guards), and may return a dict to merge into PlatformConfig.extra. It runs during load_gateway_config() after the existing generic shared-key loop and before _apply_env_overrides(), mirroring the env_enablement_fn dispatch pattern (#21306, #21331). Pure addition — no behavior change for existing platforms. Each of the eight platforms with hardcoded YAML→env blocks today (discord, telegram, whatsapp, slack, dingtalk, mattermost, matrix, feishu, ~252 LOC in gateway/config.py) can migrate in independent follow-up PRs; the hardcoded blocks remain functional in the meantime, and their `not os.getenv(...)` guards make them no-ops for any env var the hook already set. Test coverage: 10 new tests in tests/gateway/test_platform_registry.py covering field default, callable acceptance, env mutation, extras merge, both signature args, exception swallowing, missing/non-dict sections, and env > YAML precedence. Refs #3823, #24356. Closes #24836.

Lets platform plugins own their YAML→env config bridge instead of forcing core gateway/config.py to know every platform's schema. The hook receives the full parsed config.yaml and the platform's own sub-dict, may mutate os.environ (env > YAML precedence preserved via the standard `not os.getenv(...)` guards), and may return a dict to merge into PlatformConfig.extra. It runs during load_gateway_config() after the existing generic shared-key loop and before _apply_env_overrides(), mirroring the env_enablement_fn dispatch pattern (NousResearch#21306, NousResearch#21331). Pure addition — no behavior change for existing platforms. Each of the eight platforms with hardcoded YAML→env blocks today (discord, telegram, whatsapp, slack, dingtalk, mattermost, matrix, feishu, ~252 LOC in gateway/config.py) can migrate in independent follow-up PRs; the hardcoded blocks remain functional in the meantime, and their `not os.getenv(...)` guards make them no-ops for any env var the hook already set. Test coverage: 10 new tests in tests/gateway/test_platform_registry.py covering field default, callable acceptance, env mutation, extras merge, both signature args, exception swallowing, missing/non-dict sections, and env > YAML precedence. Refs NousResearch#3823, NousResearch#24356. Closes NousResearch#24836.

…ams parity) First migration of an existing built-in platform adapter to the plugin system established by IRC / Teams / LINE / Google Chat. Closes NousResearch#24325; advances the umbrella refactor in NousResearch#3823. Matches Teams' shape exactly — adapter under ``plugins/platforms/discord/`` with the standard ``__init__.py`` / ``adapter.py`` / ``plugin.yaml`` shell, ``register(ctx)`` entry point, **no back-compat shim** at the old import path, and full parity for the four hooks Teams uses plus the ``apply_yaml_config_fn`` hook that landed in NousResearch#25443 (the Discord plugin is the first consumer of that hook): * ``standalone_sender_fn`` — out-of-process cron delivery via REST API * ``setup_fn`` — interactive ``hermes setup gateway`` wizard * ``apply_yaml_config_fn`` — translate ``config.yaml`` ``discord:`` keys into ``DISCORD_*`` env vars (replaces the hardcoded block in ``gateway/config.py``) * ``is_connected`` — declares connection state from ``DISCORD_BOT_TOKEN`` * ``check_fn`` — lazy-installs ``discord.py`` on demand * plus ``allowed_users_env``, ``allow_all_env``, ``cron_deliver_env_var``, ``max_message_length``, ``emoji``, ``required_env``, ``install_hint`` * ``gateway/platforms/discord.py`` (5,101 LOC) → ``plugins/platforms/discord/adapter.py`` (git rename, R090). * New ``plugins/platforms/discord/{__init__.py, plugin.yaml}`` with ``requires_env`` / ``optional_env`` declarations. * Append ``register(ctx)`` block + new hook implementations (``_standalone_send``, ``interactive_setup``, ``_apply_yaml_config``, ``_clean_discord_user_ids``, ``_is_connected``, ``_build_adapter``, plus helpers ``_DISCORD_CHANNEL_TYPE_PROBE_CACHE`` etc.) to the adapter. * Replace the ``Platform.DISCORD elif`` branch in ``GatewayRunner._create_adapter()`` (−9 LOC) with a generic post-creation hook (+6 LOC) in the registry path: any plugin adapter that declares a ``gateway_runner`` attribute now gets it auto-injected. Webhook's built-in branch is unchanged (it doesn't go through the registry path). * Move ``_send_discord`` (190 LOC) and helpers (``_DISCORD_CHANNEL_TYPE_PROBE_CACHE``, ``_remember_channel_is_forum``, ``_probe_is_forum_cached``, ``_derive_forum_thread_name``) from ``tools/send_message_tool.py`` into the plugin as ``_standalone_send``. * Wire via ``standalone_sender_fn=_standalone_send`` (Teams pattern; same gap fixed in NousResearch#21804 for other plugin platforms). * Replace the Discord ``elif`` in ``tools/send_message_tool.py`` ``_send_to_platform`` with a 10-line registry-hook dispatch. * Drop the ``DiscordAdapter`` import and the ``Platform.DISCORD: DiscordAdapter.MAX_MESSAGE_LENGTH`` ``_MAX_LENGTHS`` entry — the registry's ``max_message_length=2000`` covers it. * Move ``_setup_discord`` and ``_clean_discord_user_ids`` (68 LOC) from ``hermes_cli/setup.py`` into the plugin as ``interactive_setup``. * Wire via ``setup_fn=interactive_setup``. CLI helpers (``prompt``, ``print_info``, etc.) are lazy-imported so the plugin's module-load surface stays minimal. * Remove ``"discord": _s._setup_discord`` from ``hermes_cli/gateway.py::_builtin_setup_fn``. * Remove the entire 32-line ``_PLATFORMS["discord"]`` static dict entry — Discord's setup metadata is now discovered dynamically via ``_all_platforms()`` from the registry entry. * Move the 59-line ``discord_cfg`` YAML→env bridge from ``gateway/config.py::load_gateway_config()`` into the plugin as ``_apply_yaml_config``. Covers ``require_mention``, ``thread_require_mention``, ``free_response_channels``, ``auto_thread``, ``reactions``, ``ignored_channels``, ``allowed_channels``, ``no_thread_channels``, ``allow_mentions.{everyone,roles,users, replied_user}``, and ``reply_to_mode`` (including the YAML 1.1 ``off``-as-False coercion and the ``extra.reply_to_mode`` fallback). * Wire via ``apply_yaml_config_fn=_apply_yaml_config``. * The hook runs BEFORE ``_apply_env_overrides`` and after the generic shared-key loop, exactly as documented in ``website/docs/developer-guide/adding-platform-adapters.md``. * Behavior is preserved exactly — every assignment still uses ``not os.getenv(...)`` guards so env vars take precedence over YAML. All 78 references to the old import path are rewritten — no back-compat shim: * 51 ``from gateway.platforms.discord import X`` → ``from plugins.platforms.discord.adapter import X`` * 5 ``import gateway.platforms.discord as discord_platform`` → ``import plugins.platforms.discord.adapter as discord_platform`` * 1 ``from gateway.platforms import discord as discord_mod`` → ``from plugins.platforms.discord import adapter as discord_mod`` * 21 ``mock.patch("gateway.platforms.discord.X")`` strings → ``mock.patch("plugins.platforms.discord.adapter.X")`` * 1 docstring reference in ``hermes_cli/commands.py`` * 1 import in ``tools/send_message_tool.py`` (now removed entirely) The import-safety test in ``tests/gateway/test_discord_imports.py`` is updated to purge the new canonical module name from ``sys.modules``. **38 files changed, +621 / −473** — net positive due to the YAML hook implementation (89 new LOC in the plugin trading for 59 deleted in core), but every line moved has a clear plugin home now. The git rename is detected at R090 because the adapter gained ~340 LOC of moved-in hook implementations (``_standalone_send`` + ``interactive_setup`` + ``_apply_yaml_config`` + helpers). * All 568 Discord-specific tests pass across 25 ``test_discord_*.py`` files plus voice/send/text-batching/reload-skills/stream-consumer/ integration tests. * All 147 tests in the YAML-touching subset (``test_discord_reply_mode``, ``test_discord_free_response``, ``test_discord_allowed_channels``, ``test_discord_allowed_mentions``, ``test_discord_channel_controls``, ``test_discord_reactions``, ``test_discord_thread_persistence``, ``test_runtime_footer``) pass — this is the strongest signal that the YAML→env hook behaves identically to the legacy block. * Broader gateway/cron/integration sweep (1297 tests) introduces zero new failures vs ``main``. Pre-existing failures in ``tests/gateway/test_tts_media_routing.py`` and ``tests/e2e/test_platform_commands.py`` reproduce identically on the unchanged ``main`` revision. * Plugin discovery sanity check confirms Discord registers alongside the other four platform plugins: Registered platforms: ['discord', 'google_chat', 'irc', 'line', 'teams'] These Discord-shaped tendrils in core were **deliberately not moved** — they are generic platform-registry concerns affecting every platform, not Discord-specific: * ``gateway/config.py:1205`` ``DISCORD_BOT_TOKEN → config.token`` env enablement — same shape Telegram has. The existing ``env_enablement_fn`` registry hook only seeds ``extra``, not ``.token``, so it can't replace this without an adapter refactor to read from ``extra["bot_token"]``. * ``gateway/run.py`` voice-mode hooks (``self.adapters.get(Platform.DISCORD)`` for ``start_voice_mode``/``stop_voice_mode``), role-based auth, ``DISCORD_ALLOW_BOTS`` branch in ``_is_user_authorized``, ``_UPDATE_ALLOWED_PLATFORMS`` frozenset, and the per-platform allowlist maps — generic platform-registry concerns. * ``Platform.DISCORD`` enum literal — stable identifier used as dict keys throughout the codebase; removing it is a separate refactor with no real benefit. * ``tools/discord_tool.py`` and ``tools/environments/local.py`` — first-class agent tools and env-passthrough config, neither is the gateway adapter. Each of these is worth its own scoping issue when the time comes.

Lets platform plugins own their YAML→env config bridge instead of forcing core gateway/config.py to know every platform's schema. The hook receives the full parsed config.yaml and the platform's own sub-dict, may mutate os.environ (env > YAML precedence preserved via the standard `not os.getenv(...)` guards), and may return a dict to merge into PlatformConfig.extra. It runs during load_gateway_config() after the existing generic shared-key loop and before _apply_env_overrides(), mirroring the env_enablement_fn dispatch pattern (NousResearch#21306, NousResearch#21331). Pure addition — no behavior change for existing platforms. Each of the eight platforms with hardcoded YAML→env blocks today (discord, telegram, whatsapp, slack, dingtalk, mattermost, matrix, feishu, ~252 LOC in gateway/config.py) can migrate in independent follow-up PRs; the hardcoded blocks remain functional in the meantime, and their `not os.getenv(...)` guards make them no-ops for any env var the hook already set. Test coverage: 10 new tests in tests/gateway/test_platform_registry.py covering field default, callable acceptance, env mutation, extras merge, both signature args, exception swallowing, missing/non-dict sections, and env > YAML precedence. Refs NousResearch#3823, NousResearch#24356. Closes NousResearch#24836.

…ed=False, not silently fall back to check_fn Two bugs surfaced by PR #24356 migrating Discord into the registry: 1. plugins/platforms/discord/adapter.py::_is_connected — read DISCORD_BOT_TOKEN via hermes_cli.gateway.get_env_value (the abstraction tests patch) instead of os.getenv directly. The legacy non-registry path used get_env_value; bypassing it broke test_setup_openclaw_migration which patches gateway_mod.get_env_value to simulate a hermetic env. 2. hermes_cli/gateway.py::_platform_status — when entry.is_connected is defined and returns False, return 'not configured' immediately. Don't fall back to entry.check_fn(), which would let 'SDK is installed' override 'no token configured' and incorrectly report the platform as ready. The fallback to check_fn is the right behaviour only when is_connected is None (not registered). Fixes 5 test failures observed on CI for PR #24356: - tests/hermes_cli/test_setup.py::test_setup_gateway_skips_service_install_when_systemctl_missing - tests/hermes_cli/test_setup.py::test_setup_gateway_in_container_shows_docker_guidance - tests/hermes_cli/test_setup_irc.py::TestIRCGatewaySetupFreshInstall::test_setup_gateway_irc_counts_as_messaging_platform - tests/hermes_cli/test_setup_openclaw_migration.py::TestGetSectionConfigSummary::test_gateway_returns_none_without_tokens - tests/hermes_cli/test_setup_openclaw_migration.py::TestSetupWizardSkipsConfiguredSections::test_sections_skipped_when_migration_imported_settings Same _platform_status bug exists for sibling plugin platforms (teams, google_chat) whose check_fn returns true on SDK install alone; their tests just never exercised the registry path before. The bug only became test-visible when Discord migrated into the registry. Validation: 11,167 tests across tests/gateway/ + tests/cron/ + tests/tools/test_send_message_tool.py + tests/hermes_cli/ pass with zero failures.

teknium1 · 2026-05-22T21:21:52Z

Salvaged via #30591 (merged). Your migration commit was cherry-picked onto current main with your authorship preserved in git log. Three small conflicts resolved against the 840-commit drift on main (typing 429 graceful, voice note transcription, lazy view-class init, ruff PLR6201 sweep, allow_any_attachment all flowed in via auto-merge). One follow-up commit fixed a latent _platform_status fallback bug that the migration exposed — is_connected=False was silently overridden by check_fn returning true on installed SDK. Thanks for the migration!

…ed=False, not silently fall back to check_fn Two bugs surfaced by PR NousResearch#24356 migrating Discord into the registry: 1. plugins/platforms/discord/adapter.py::_is_connected — read DISCORD_BOT_TOKEN via hermes_cli.gateway.get_env_value (the abstraction tests patch) instead of os.getenv directly. The legacy non-registry path used get_env_value; bypassing it broke test_setup_openclaw_migration which patches gateway_mod.get_env_value to simulate a hermetic env. 2. hermes_cli/gateway.py::_platform_status — when entry.is_connected is defined and returns False, return 'not configured' immediately. Don't fall back to entry.check_fn(), which would let 'SDK is installed' override 'no token configured' and incorrectly report the platform as ready. The fallback to check_fn is the right behaviour only when is_connected is None (not registered). Fixes 5 test failures observed on CI for PR NousResearch#24356: - tests/hermes_cli/test_setup.py::test_setup_gateway_skips_service_install_when_systemctl_missing - tests/hermes_cli/test_setup.py::test_setup_gateway_in_container_shows_docker_guidance - tests/hermes_cli/test_setup_irc.py::TestIRCGatewaySetupFreshInstall::test_setup_gateway_irc_counts_as_messaging_platform - tests/hermes_cli/test_setup_openclaw_migration.py::TestGetSectionConfigSummary::test_gateway_returns_none_without_tokens - tests/hermes_cli/test_setup_openclaw_migration.py::TestSetupWizardSkipsConfiguredSections::test_sections_skipped_when_migration_imported_settings Same _platform_status bug exists for sibling plugin platforms (teams, google_chat) whose check_fn returns true on SDK install alone; their tests just never exercised the registry path before. The bug only became test-visible when Discord migrated into the registry. Validation: 11,167 tests across tests/gateway/ + tests/cron/ + tests/tools/test_send_message_tool.py + tests/hermes_cli/ pass with zero failures.

…ed=False, not silently fall back to check_fn Two bugs surfaced by PR NousResearch#24356 migrating Discord into the registry: 1. plugins/platforms/discord/adapter.py::_is_connected — read DISCORD_BOT_TOKEN via hermes_cli.gateway.get_env_value (the abstraction tests patch) instead of os.getenv directly. The legacy non-registry path used get_env_value; bypassing it broke test_setup_openclaw_migration which patches gateway_mod.get_env_value to simulate a hermetic env. 2. hermes_cli/gateway.py::_platform_status — when entry.is_connected is defined and returns False, return 'not configured' immediately. Don't fall back to entry.check_fn(), which would let 'SDK is installed' override 'no token configured' and incorrectly report the platform as ready. The fallback to check_fn is the right behaviour only when is_connected is None (not registered). Fixes 5 test failures observed on CI for PR NousResearch#24356: - tests/hermes_cli/test_setup.py::test_setup_gateway_skips_service_install_when_systemctl_missing - tests/hermes_cli/test_setup.py::test_setup_gateway_in_container_shows_docker_guidance - tests/hermes_cli/test_setup_irc.py::TestIRCGatewaySetupFreshInstall::test_setup_gateway_irc_counts_as_messaging_platform - tests/hermes_cli/test_setup_openclaw_migration.py::TestGetSectionConfigSummary::test_gateway_returns_none_without_tokens - tests/hermes_cli/test_setup_openclaw_migration.py::TestSetupWizardSkipsConfiguredSections::test_sections_skipped_when_migration_imported_settings Same _platform_status bug exists for sibling plugin platforms (teams, google_chat) whose check_fn returns true on SDK install alone; their tests just never exercised the registry path before. The bug only became test-visible when Discord migrated into the registry. Validation: 11,167 tests across tests/gateway/ + tests/cron/ + tests/tools/test_send_message_tool.py + tests/hermes_cli/ pass with zero failures. #AI commit#

…ed=False, not silently fall back to check_fn Two bugs surfaced by PR NousResearch#24356 migrating Discord into the registry: 1. plugins/platforms/discord/adapter.py::_is_connected — read DISCORD_BOT_TOKEN via hermes_cli.gateway.get_env_value (the abstraction tests patch) instead of os.getenv directly. The legacy non-registry path used get_env_value; bypassing it broke test_setup_openclaw_migration which patches gateway_mod.get_env_value to simulate a hermetic env. 2. hermes_cli/gateway.py::_platform_status — when entry.is_connected is defined and returns False, return 'not configured' immediately. Don't fall back to entry.check_fn(), which would let 'SDK is installed' override 'no token configured' and incorrectly report the platform as ready. The fallback to check_fn is the right behaviour only when is_connected is None (not registered). Fixes 5 test failures observed on CI for PR NousResearch#24356: - tests/hermes_cli/test_setup.py::test_setup_gateway_skips_service_install_when_systemctl_missing - tests/hermes_cli/test_setup.py::test_setup_gateway_in_container_shows_docker_guidance - tests/hermes_cli/test_setup_irc.py::TestIRCGatewaySetupFreshInstall::test_setup_gateway_irc_counts_as_messaging_platform - tests/hermes_cli/test_setup_openclaw_migration.py::TestGetSectionConfigSummary::test_gateway_returns_none_without_tokens - tests/hermes_cli/test_setup_openclaw_migration.py::TestSetupWizardSkipsConfiguredSections::test_sections_skipped_when_migration_imported_settings Same _platform_status bug exists for sibling plugin platforms (teams, google_chat) whose check_fn returns true on SDK install alone; their tests just never exercised the registry path before. The bug only became test-visible when Discord migrated into the registry. Validation: 11,167 tests across tests/gateway/ + tests/cron/ + tests/tools/test_send_message_tool.py + tests/hermes_cli/ pass with zero failures.

Lets platform plugins own their YAML→env config bridge instead of forcing core gateway/config.py to know every platform's schema. The hook receives the full parsed config.yaml and the platform's own sub-dict, may mutate os.environ (env > YAML precedence preserved via the standard `not os.getenv(...)` guards), and may return a dict to merge into PlatformConfig.extra. It runs during load_gateway_config() after the existing generic shared-key loop and before _apply_env_overrides(), mirroring the env_enablement_fn dispatch pattern (NousResearch#21306, NousResearch#21331). Pure addition — no behavior change for existing platforms. Each of the eight platforms with hardcoded YAML→env blocks today (discord, telegram, whatsapp, slack, dingtalk, mattermost, matrix, feishu, ~252 LOC in gateway/config.py) can migrate in independent follow-up PRs; the hardcoded blocks remain functional in the meantime, and their `not os.getenv(...)` guards make them no-ops for any env var the hook already set. Test coverage: 10 new tests in tests/gateway/test_platform_registry.py covering field default, callable acceptance, env mutation, extras merge, both signature args, exception swallowing, missing/non-dict sections, and env > YAML precedence. Refs NousResearch#3823, NousResearch#24356. Closes NousResearch#24836.

…ed=False, not silently fall back to check_fn Two bugs surfaced by PR NousResearch#24356 migrating Discord into the registry: 1. plugins/platforms/discord/adapter.py::_is_connected — read DISCORD_BOT_TOKEN via hermes_cli.gateway.get_env_value (the abstraction tests patch) instead of os.getenv directly. The legacy non-registry path used get_env_value; bypassing it broke test_setup_openclaw_migration which patches gateway_mod.get_env_value to simulate a hermetic env. 2. hermes_cli/gateway.py::_platform_status — when entry.is_connected is defined and returns False, return 'not configured' immediately. Don't fall back to entry.check_fn(), which would let 'SDK is installed' override 'no token configured' and incorrectly report the platform as ready. The fallback to check_fn is the right behaviour only when is_connected is None (not registered). Fixes 5 test failures observed on CI for PR NousResearch#24356: - tests/hermes_cli/test_setup.py::test_setup_gateway_skips_service_install_when_systemctl_missing - tests/hermes_cli/test_setup.py::test_setup_gateway_in_container_shows_docker_guidance - tests/hermes_cli/test_setup_irc.py::TestIRCGatewaySetupFreshInstall::test_setup_gateway_irc_counts_as_messaging_platform - tests/hermes_cli/test_setup_openclaw_migration.py::TestGetSectionConfigSummary::test_gateway_returns_none_without_tokens - tests/hermes_cli/test_setup_openclaw_migration.py::TestSetupWizardSkipsConfiguredSections::test_sections_skipped_when_migration_imported_settings Same _platform_status bug exists for sibling plugin platforms (teams, google_chat) whose check_fn returns true on SDK install alone; their tests just never exercised the registry path before. The bug only became test-visible when Discord migrated into the registry. Validation: 11,167 tests across tests/gateway/ + tests/cron/ + tests/tools/test_send_message_tool.py + tests/hermes_cli/ pass with zero failures.

Move gateway/platforms/slack.py into plugins/platforms/slack/ following the Discord (#24356) and Home Assistant (#40709) migrations. Advances #41112 / hardcoded Platform.SLACK touchpoints in core. - Adapter file renamed via git mv (history preserved). - register() exposes the platform via ctx.register_platform() instead of the Platform.SLACK elif in gateway/run.py::_create_adapter(). - _standalone_send() replaces the legacy _send_slack() helper in tools/send_message_tool.py; out-of-process cron delivery (deliver=slack) now flows through the registry's standalone_sender_fn. mrkdwn formatting moved into the plugin (was applied in _send_to_platform before chunking). - _apply_yaml_config() owns the config.yaml slack: -> SLACK_* env bridge (require_mention, strict_mention, allow_bots, free_response_channels, reactions, allowed_channels), replacing the hardcoded block in gateway/config.py. - interactive_setup() replaces hermes_cli/setup.py::_setup_slack + _write_slack_manifest_and_instruct and the static _PLATFORMS["slack"] dict in hermes_cli/gateway.py; setup metadata is discovered dynamically. - is_connected() probes SLACK_BOT_TOKEN via hermes_cli.gateway.get_env_value. - max_message_length=39000 on the PlatformEntry; the registry fallback in send_message_tool covers it (dropped the _MAX_LENGTHS entry). The SLACK_BOT_TOKEN/SLACK_HOME_CHANNEL env->PlatformConfig seeding and the _is_user_authorized allowlist maps stay in core (same as Discord/HA/Mattermost). Bug fixed during migration: the registry-driven plugin-enable pass in _apply_env_overrides re-enabled any plugin platform whose is_connected() passed, ignoring an explicit enabled: false. Slack is the first plugin with an enabled-false-wins test, so it exposed this latent bug (Discord had no such test). Added an explicit-disable guard (_enabled_explicit + enabled=False -> skip) and changed the slack env-block to read the flag instead of popping it so the guard can see it; the flag is still cleared in the final per-platform cleanup. Restores test_explicit_{top_level,platforms}_slack_enabled_false_wins. Test imports rewritten across 11 files (gateway.platforms.slack -> plugins.platforms.slack.adapter). The _setup_slack home-channel tests moved to tests/gateway/test_slack_plugin_setup.py exercising interactive_setup. The test_send_message_tool slack-formatting tests now patch the registry standalone_sender_fn (via _patch_slack_standalone_sender) and assert the mrkdwn-formatted text reaches the wire. Validation: 706 targeted tests pass (slack/config/setup/registry/send/media suites); 18/18 live E2E checks pass (real plugin discovery + registry resolves SlackAdapter, env-only enable, standalone sender wired, YAML bridge, dynamic setup discovery).

alt-glitch added type/refactor Code restructuring, no behavior change P3 Low — cosmetic, nice to have comp/gateway Gateway runner, session dispatch, delivery comp/plugins Plugin system and bundled plugins platform/discord Discord bot adapter labels May 12, 2026

kshitijk4poor force-pushed the spike/discord-plugin-extraction branch 2 times, most recently from 55b485b to 45ce4f7 Compare May 12, 2026 14:09

This was referenced May 13, 2026

refactor(plugins): add apply_yaml_config_fn registry hook to let plugins own their YAML→env config bridges #24836

Closed

refactor(plugins): add apply_yaml_config_fn registry hook for YAML→env config bridges #24849

Closed

teknium1 mentioned this pull request May 14, 2026

refactor(plugins): add apply_yaml_config_fn registry hook for YAML→env config bridges #25443

Merged

kshitijk4poor force-pushed the spike/discord-plugin-extraction branch from 45ce4f7 to 87d1697 Compare May 14, 2026 08:53

kshitijk4poor force-pushed the spike/discord-plugin-extraction branch from 87d1697 to 29bc396 Compare May 15, 2026 07:49

teknium1 mentioned this pull request May 22, 2026

refactor(gateway): migrate Discord adapter to bundled plugin (salvage of #24356) #30591

Merged

teknium1 closed this May 22, 2026

kshitijk4poor mentioned this pull request May 26, 2026

chore(gateway): drop plugin-migrated platforms from /update allowlist #32525

Closed

teknium1 mentioned this pull request Jun 6, 2026

chore(gateway): drop plugin-migrated platforms from /update allowlist #40711

Merged

kshitijk4poor mentioned this pull request Jun 7, 2026

tracking: gateway platform → bundled-plugin migration #41112

Open

28 tasks

teknium1 mentioned this pull request Jun 7, 2026

refactor(gateway): migrate Slack adapter to bundled plugin #41160

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

refactor(gateway): migrate Discord adapter to bundled plugin#24356

refactor(gateway): migrate Discord adapter to bundled plugin#24356
kshitijk4poor wants to merge 1 commit into
NousResearch:mainfrom
kshitijk4poor:spike/discord-plugin-extraction

kshitijk4poor commented May 12, 2026 •

edited

Loading

Uh oh!

teknium1 commented May 22, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Conversation

kshitijk4poor commented May 12, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Why Discord first

Hooks wired

What changed

Diff stats

Test coverage

Out of scope (separate concerns, not Discord-specific)

Uh oh!

teknium1 commented May 22, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

kshitijk4poor commented May 12, 2026 •

edited

Loading