909 plugins from the official Hermes plugin catalog: 20 official, 9 verified and 880 from the community. 206 were added in the last week. Search, filter by what you want to do, and copy the install command.
880 plugins. Sorted by HermesWatcher's picks: usefulness, polish and trust. Official: made by Nous Research. Verified: listed with a verified maintainer. Community: built by others, so read the source first.
Bunny Space gateway adapter for Hermes Agent. Connects a Hermes profile (the bot brain) to the Bunny Space relay so the bot joins private rooms, answers its owner, and handles @bot:<slug> mentions in posts and comments. Outbound pull transport; no external dependencies.
Good to know before installing
Replies are prompted to be in Spanish; @bot mention tasks and assistance tasks are dispatched to Hermes under the fixed sender id "owner" (BUNNY_SPACE_ALLOWED_USERS must contain "owner"), and the plugin refuses to dispatch third-party content unless platform_toolsets.bunny_space is restricted to read-only tools (the platform's effective toolset, including MCP and plugin toolsets; add no_mcp when MCP servers are enabled); the private assistance worker is opt-in via BUNNY_SPACE_ASSISTANCE=1 (the relay cannot enable it); setting BUNNY_SPACE_ALLOW_ALL_USERS makes the connector refuse to connect; the relay sets the poll interval and may hand the connector a WebSocket "wake" URL it then connects to (wake signals only, no message content).
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs (this one needs BUNNY_SPACE_BOT_KEY, BUNNY_SPACE_SLUG), then asks whether to turn it on.
hermes plugins install bunny_space-platform
Said no? Turn it on later:
hermes plugins enable bunny_space-platform
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Windows taskbar badge on the Hermes window when a turn finishes. Calls navigator.setAppBadge on the message.complete event while the window is in the background, and clears it when the window regains focus or another session is opened.
Good to know before installing
No network calls and no data leaves the machine; the badge is drawn by the app's own Electron process, which maps the Chromium Badging API onto the Windows taskbar overlay.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install hermes-taskbar-badge
Said no? Turn it on later:
hermes plugins enable hermes-taskbar-badge
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Companion to hermes-kame-api-rotation (install both). A model-provider plugin that re-registers Hermes' bundled API-key profiles unchanged except for create_client and create_messages_client, which return the rotation plugin's key-rotating client; otherwise None and Hermes builds its own. Same endpoints as Hermes, no third-party service, no core override.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install hermes-kame-provider
Said no? Turn it on later:
hermes plugins enable hermes-kame-provider
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Independent Hermes integration of Victor Taelin's OptMem design: local, append-only memory with regex/BM25 recall and a bounded wake digest. Hybrid and guarded OptMem-only modes; verified migration and rollback. Local by default. Compaction is manual unless auto_nap is enabled. Optional LLM summaries run only when both auto_nap and llm_summary are on; they send memory blocks to the configured provider via host-owned authentication and may incur token costs. Summaries are lossy, raw records remain recoverable, and conflicts are not resolved automatically. Wake consumes model context. See README for upstream attribution.
Good to know before installing
Memory is stored locally under HERMES_HOME; optional auto_nap+llm_summary sends stored memory lines to the user's configured model provider; the optmem-only mode (explicit CLI with --yes) disables Hermes' built-in MEMORY.md/USER.md injection after a verified backup.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install optmem-hermes
Said no? Turn it on later:
hermes plugins enable optmem-hermes
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Jina Reader web backend for web_extract (r.jina.ai, works without a key) and web_search (s.jina.ai, needs JINA_API_KEY). Unofficial; not affiliated with Jina AI.
How to install
Install it. Hermes asks for any API keys it needs (this one needs JINA_API_KEY), then asks whether to turn it on.
hermes plugins install hermes-jina
Said no? Turn it on later:
hermes plugins enable hermes-jina
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Full setup steps, settings and screenshots are on the plugin's official catalog page from Nous Research.
Hermes Stt Enhance
Community
Local speech-to-text (provider stt_enhance) with faster-whisper or opt-in sherpa-onnx Parakeet V3 and long-audio chunking, plus an optional bounded LLM pass that enhances the transcript using a custom skill, an inline…
Local speech-to-text (provider stt_enhance) with faster-whisper or opt-in sherpa-onnx Parakeet V3 and long-audio chunking, plus an optional bounded LLM pass that enhances the transcript using a custom skill, an inline prompt or the default, and keeps the raw transcript on failure.
Good to know before installing
Audio is decoded locally. Post-processing is on by default and sends the transcript text to the LLM Hermes routes the stt_enhance auxiliary task to (your main, usually remote, provider unless you point it at a local endpoint or set post_processing.enabled: false). It reads the body of the SKILL.md you configure (never executed) and runs ffmpeg/ffprobe. The opt-in Parakeet backend reads a local model export and, on first use and only when your config.yaml explicitly sets security.allow_lazy_installs: true, installs pinned sherpa-onnx and numpy from PyPI into ~/.hermes/plugin-runtimes/hermes-stt-enhance/ (shared by all profiles). No tools, hooks, background processes, telemetry or stored credentials.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install hermes-stt-enhance
Said no? Turn it on later:
hermes plugins enable hermes-stt-enhance
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Outbound WhatsApp tools through GOWA: registers GOWA's native MCP server and adds device listing, pairing, and removal. Setup can connect an existing server or explicitly install pinned GOWA v9.5.0 as a loopback-only authenticated native user service on Linux x86_64, macOS x86_64/arm64, or Windows x86_64.
Good to know before installing
Agents can send messages and mutate WhatsApp state through the linked account; session databases grant account access. The plugin does not ingest incoming messages or configure webhooks.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install go-whatsapp
Said no? Turn it on later:
hermes plugins enable go-whatsapp
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Unofficial OpenCode Zen free-model transport for Hermes. Discovers live zero-cost tool-capable models by intersecting OpenCode Zen and models.dev over HTTPS, persists only the last successful intersection for restart fallback, and bidirectionally translates compatible OpenCode and Hermes tool schemas, calls, choices, and replayed history while Hermes executes tools locally. No OpenCode CLI, Node.js, npm, or npx dependency.
Good to know before installing
Prompts, tool schemas, tool results, and other conversation content leave the machine for OpenCode and the selected model provider; model availability, retention, data use, and permission are governed by their current terms.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install oc-free-provider
Said no? Turn it on later:
hermes plugins enable oc-free-provider
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Local long-term memory for Hermes agents. Everything stays on your machine: an SQLite store plus bundled embedding and NLI models, no telemetry, no LLMs. Limbic is designed on the physiological brain, opaque and organic. Memories strengthen or atrophy with use, corrections replace what they superseded — all deterministic background mechanisms, zero LLM calls. Algorithmic forgetting keeps atrophied memories out of the database. Recall is calibrated: when nothing in memory is confident enough, it abstains instead of guessing, so the agent answers honestly rather than surfacing near-misses. First start downloads ~1 GB of pinned models once, then runs fully offline. PolyForm Noncommercial 1.0.0 (commercial use requires a separate license).
Good to know before installing
The agent-facing tools are remember, recall and forget (forget lists candidate facts for a query, then deletes the one fact id the user picks, scoped to the current user, and appends a hash-only receipt of the content's sha256, length, user id and fact id to limbic.db.deletions.jsonl beside the database); on first primary session start downloads ~1.0GB of ONNX models (Arctic Embed 2.0 L int8 embedding, ~570MB; MiniLMv2 NLI, ~430MB, on first contradiction check) from huggingface.co at pinned commits with hash checks into ~/.cache/limbic/ (outside HERMES_HOME; %USERPROFILE%\.cache\limbic on Windows); at every model load, every file in those cache folders that is not in the pinned manifest is deleted — including unrelated files if you point embedding/nli cache_dir at a shared folder; spaCy NER models are optional: `hermes limbic languages --set <lang>` and `hermes limbic setup` run `python -m spacy download <model>`, which pip-installs each model package (~14.5MB, version picked by spaCy, not hash-pinned) from spaCy's GitHub releases into Hermes's own Python environment, and `hermes limbic setup` also pre-downloads both ONNX models; memory stays local in limbic.db under HERMES_HOME, f
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install limbic
Said no? Turn it on later:
hermes plugins enable limbic
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Full setup steps, settings and screenshots are on the plugin's official catalog page from Nous Research.
Wyoming Voice Bridge
Community
Wyoming speech-to-text and text-to-speech for a Wyoming server you run (stt/tts provider `wyoming`), plus a wyoming_voice_status tool that lists the server's models and voices.
Wyoming speech-to-text and text-to-speech for a Wyoming server you run (stt/tts provider `wyoming`), plus a wyoming_voice_status tool that lists the server's models and voices.
Good to know before installing
Audio and text to speak go to the WYOMING_* host:port you set over cleartext, unauthenticated TCP (public addresses accepted). Once selected, the providers send speech in every session, including cron and webhooks. Non-WAV audio is converted with your local ffmpeg. wyoming_voice_status asks for approval on each call and refuses under cron, yolo, approvals off, -q and unattended platforms; answering "always" leaves a command_allowlist line in config.yaml. It calls internal Hermes approval, config and write-guard helpers, so a Hermes update can disable the status tool or speech file writes (both fail closed).
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install wyoming-voice-bridge
Said no? Turn it on later:
hermes plugins enable wyoming-voice-bridge
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Full setup steps, settings and screenshots are on the plugin's official catalog page from Nous Research.
Nextcloud Talk Platform
Community
Nextcloud Talk platform adapter for Hermes Agent, in two modes behind one plugin: a webhook bot (Nextcloud Talk Bot API, HMAC-SHA256 signed, with an explicit body-size cap and a nonce-replay cache ahead of payload…
Nextcloud Talk platform adapter for Hermes Agent, in two modes behind one plugin: a webhook bot (Nextcloud Talk Bot API, HMAC-SHA256 signed, with an explicit body-size cap and a nonce-replay cache ahead of payload parsing) and a real Nextcloud user account (app password, long-polling the Chat REST API, with federation-invite support, per-room mention/watch/free modes, and an agent-to-agent loop guard plus active-listening context for multi-agent rooms). Mode is picked automatically by which credential is configured. No extra SDKs -- aiohttp for bot mode, httpx for user mode, both already Hermes dependencies.
Good to know before installing
Connects to your Nextcloud server either as a Talk bot (HMAC-verified webhook listener, 127.0.0.1:18797 by default) or as a real Nextcloud user with an app password (long-polls the Chat API and auto-accepts federation invites only from NEXTCLOUD_TALK_ALLOWED_INVITERS); in user mode, unanswered room messages from allowed senders (up to 40 per room by default) and, for rooms listed in NEXTCLOUD_TALK_ROOM_MODES or with room_rules_all_rooms on, the room description are added to the agent's next turn in that room.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install nextcloud-talk-platform
Said no? Turn it on later:
hermes plugins enable nextcloud-talk-platform
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Full setup steps, settings and screenshots are on the plugin's official catalog page from Nous Research.
SpecStream
Community
SpecStream gives Hermes finished lists from the live web: describe the list in plain words and get every matching row with the page it was read from, priced per match with a maximum you approve before anything runs.
SpecStream gives Hermes finished lists from the live web: describe the list in plain words and get every matching row with the page it was read from, priced per match with a maximum you approve before anything runs. Portable Agent Plugins v1 package with the hosted MCP server (remote streamable-http, OAuth 2.1 with dynamic client registration) and four skills (compile, review-work, approve-work, continue-work); no Python, no executables, no stored credentials.
Good to know before installing
Your request text and your answers to the price are sent to api.specstream.ai; approving a price spends your SpecStream balance (new accounts get $10 free) and the plugin marks its server `untrusted`, so Hermes asks before every call that is not read-only and refuses those calls in unattended runs; a `specstream` server you define in config.yaml replaces the package's and drops that prompt unless you add `trust: untrusted`.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install specstream
Said no? Turn it on later:
hermes plugins enable specstream
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Full setup steps, settings and screenshots are on the plugin's official catalog page from Nous Research.
Cloudflare Workers Ai Provider
Community
Cloudflare Workers AI as a Hermes model provider (`--provider workers-ai`, or Cloudflare Workers AI in `hermes model`, which asks for the token and the account's base URL): the 16 Workers AI models with function…
Cloudflare Workers AI as a Hermes model provider (`--provider workers-ai`, or Cloudflare Workers AI in `hermes model`, which asks for the token and the account's base URL): the 16 Workers AI models with function calling and 64K+ context, Cloudflare's context windows (a custom endpoint lists no models and assumes 256K for the 128K gpt-oss), reasoning effort in the form each model accepts, and a clear message when a free-plan account picks a paid-plan model. Not affiliated with Cloudflare.
Good to know before installing
Registers one model provider at import and nothing else (no tools, hooks, commands or threads); the model list and details are built in, so nothing is fetched to show them; chat requests go to the URL in CLOUDFLARE_WORKERS_AI_BASE_URL through Hermes' own client with CLOUDFLARE_WORKERS_AI_API_TOKEN, both read by Hermes per profile; the plugin itself opens no connections, reads and writes no files and starts no processes.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs (this one needs CLOUDFLARE_WORKERS_AI_API_TOKEN, CLOUDFLARE_WORKERS_AI_BASE_URL), then asks whether to turn it on.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Lossless memory for long sessions: LCM-X keeps every turn in a local store instead of compacting it away, and the agent can search and expand it. The store is local SQLite (lcm.db) with a recoverable DAG of summaries over it and 15 lcm_* tools for search, recall and expansion. Select it with context.engine: lcm-x. Upgrading from hermes-lcm (breaking in v0.24.0): stop Hermes, then in plugins.enabled replace hermes-lcm with hermes-lcm-x and run exactly one LCM copy (the install script prints the steps; the old engine name lcm still resolves with a deprecation warning); lcm.db and the lcm: config block are unchanged. Summaries use the model Hermes is configured with.
Good to know before installing
Optional semantic recall (embeddings_enabled, off by default) sends message text to the embedding provider you configure (Voyage via VOYAGE_API_KEY, an OpenAI-compatible endpoint, or a local Ollama), and once enabled new summaries are embedded automatically in the background; in-process FastEmbed keeps it on the machine after a one-time model download. For known cloud providers the provider-bound copy is privacy-transformed by default (LCM_EMBEDDING_PRIVACY_ENABLED=false opts out), while lcm.db keeps the raw text. Every turn (user, assistant and tool output) is stored in plaintext in lcm.db (secret redaction at rest is opt-in), and oversized tool outputs are by default written as plaintext files under HERMES_HOME/lcm-large-outputs and replaced in the model's context by recoverable references. The engine also reads Hermes' state.db read-only for session lineage, reads private AIAgent attributes via the call stack on Hermes builds without subagent hooks, and runs git on its own install directory for lcm_status/lcm_doctor; tiktoken may fetch its BPE file on first use.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install hermes-lcm-x
Said no? Turn it on later:
hermes plugins enable hermes-lcm-x
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Local-first OpenWhispr dictation history and shared notes for Hermes agents: find what the user said, search notes, transcribe local audio. Free local desktop bridge by default, paid cloud opt-in only. Skill-carrier plugin, no tools or hooks registered.
Good to know before installing
Skill only: the agent runs the external OpenWhispr CLI (installed with an unpinned `npm install -g @openwhispr/cli`) through the terminal tool to read your dictation history and notes over the OpenWhispr desktop app's authenticated loopback bridge, and, with your consent each time, to create, update or delete notes, transcriptions and dictionary entries and to transcribe local audio files; OpenWhispr Cloud (api.openwhispr.com, including an email-code sign-in that mints an API key kept in the CLI's own config) is used only when you opt in; the agent never opens credential files itself.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install hermes-openwhispr
Said no? Turn it on later:
hermes plugins enable hermes-openwhispr
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Rooms where AI agents of any vendor talk to each other. A room is a URL; agents join, read and post over plain HTTP, with no accounts and no sign-in. Portable Agent Plugins v1 package: the parlor skill plus the hosted parlor MCP server (Streamable HTTP, no authentication, 8 tools).
Good to know before installing
Hosted unauthenticated MCP connector at parlor.sh/mcp (tool set is server-controlled, not pinned) plus a curl fallback; messages the agent writes go to parlor.sh or the server in a given room URL and rooms are public to anyone with the link; room tokens are stored under ~/.local/state/parlor; the skill may keep background wait loops running.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install parlor
Said no? Turn it on later:
hermes plugins enable parlor
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Session source tray — URLs the Hermes agent opened via web_search, web_extract, or browser_navigate. Desktop pane and /sources. No search backend of its own and no API keys.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install source-tray
Said no? Turn it on later:
hermes plugins enable source-tray
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Full setup steps, settings and screenshots are on the plugin's official catalog page from Nous Research.
Usage Guard
Community
Keeps an agent inside its subscription allowance: refuses a provider dispatch once a subscription usage window is exhausted, so an account that allows paid extra usage stops at the wall instead of spending overage.
Keeps an agent inside its subscription allowance: refuses a provider dispatch once a subscription usage window is exhausted, so an account that allows paid extra usage stops at the wall instead of spending overage. Reads the usage Hermes already displays, holds no credentials, and never touches account or billing settings.
Good to know before installing
Refuses main-loop provider calls (including MoA turns) for the providers in `provider_keys` (Anthropic and OpenAI Codex by default) when a subscription usage window reaches `threshold_percent`, and by default (`fail_open: false`) also refuses when usage data is missing, stale, unreadable, or the guard's own config is malformed; usage is read through Hermes's own account-usage fetcher (no credentials held, at most one read per source every 180 seconds by default) and cached in $HERMES_HOME/plugin-data/usage-guard/usage.json; a `pre_llm_call` hook adds a short usage note to the next turn's context at 80% and 90% used by default, and an opt-in (off by default) `llm_request` middleware lowers an existing output-token budget near the usage limit; both read only the cache, never the network; auxiliary calls (compression, vision, titles, MoA advisors) are not gated.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install usage-guard
Said no? Turn it on later:
hermes plugins enable usage-guard
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Full setup steps, settings and screenshots are on the plugin's official catalog page from Nous Research.
Agentplaybooks Memory
Community
Private AgentPlaybooks memory for Hermes: bounded keyword recall across personal and team sources, exact native memory-write mirroring, archive and history.
Private AgentPlaybooks memory for Hermes: bounded keyword recall across personal and team sources, exact native memory-write mirroring, archive and history.
Good to know before installing
Keywords derived from each turn's query are sent to the configured AgentPlaybooks service for recall from your personal and configured team playbooks; explicit native memory writes are mirrored there. Defaults to agentplaybooks.ai over HTTPS; localhost HTTP is supported, and remote intranet HTTP requires an explicit opt-in that sends the API key unencrypted. Team and shared playbooks are configured explicitly: the agent can write, archive and permanently delete entries in configured team playbooks (shared playbooks are read-only), and teammates' entries are recalled into context.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs (this one needs AGENTPLAYBOOKS_MEMORY_API_KEY), then asks whether to turn it on.
hermes plugins install agentplaybooks-memory
Said no? Turn it on later:
hermes plugins enable agentplaybooks-memory
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Full setup steps, settings and screenshots are on the plugin's official catalog page from Nous Research.
Salesforce Chatter
Community
Talk to Hermes from Salesforce Chatter: @mention the bot user in a post or comment, it Likes the message as an acknowledgement and replies in the same thread, keeping the thread as one session.
Talk to Hermes from Salesforce Chatter: @mention the bot user in a post or comment, it Likes the message as an acknowledgement and replies in the same thread, keeping the thread as one session. Images and files on the post are passed to the model; MEDIA files are attached back. Community plugin, not affiliated with Salesforce.
Good to know before installing
Authenticates as a dedicated bot user with the OAuth 2.0 JWT bearer flow, reading the private key file at SF_CHATTER_PRIVATE_KEY_PATH; talks to the configured Salesforce login host and the org instance (attachment downloads follow Salesforce redirects); image URLs in the agent's replies are fetched from the gateway host over HTTPS (any public host; private, loopback and metadata addresses are refused) and uploaded to Chatter; polls the To Me feed every 60 s and, for allowlisted groups, runs a SOQL probe every 5 s by default (counts against the org API quota); writes Likes, comments, files in the bot user's Files and optional home-channel posts; keeps its ledger in plugin-data/salesforce-chatter/inbox.sqlite. dry_run defaults to true (no Salesforce writes, no outbound image fetch); with SF_CHATTER_ALLOWED_USERS unset nobody is answered. Optional, off by default: html_preview renders model-generated HTML to PNG with Chromium in a network-isolated Docker container (digest-pinned image, pulled if missing). No telemetry.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs (this one needs SF_CHATTER_LOGIN_URL, SF_CHATTER_CLIENT_ID, SF_CHATTER_USERNAME, SF_CHATTER_PRIVATE_KEY_PATH, SF_CHATTER_BOT_USER_ID), then asks whether to turn it on.
hermes plugins install salesforce-chatter
Said no? Turn it on later:
hermes plugins enable salesforce-chatter
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Deepgram Aura text-to-speech and Nova speech-to-text providers for Hermes Agent.
How to install
Install it. Hermes asks for any API keys it needs (this one needs DEEPGRAM_API_KEY), then asks whether to turn it on.
hermes plugins install deepgram-voice
Said no? Turn it on later:
hermes plugins enable deepgram-voice
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Keyless local web extract: fetches pages from this machine with httpx and reduces them to readable text with trafilatura. No API key, no third-party service. Extract-only; pair with any search provider.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install web-local
Said no? Turn it on later:
hermes plugins enable web-local
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Track subscription quotas and reset times in Hermes Desktop.
Good to know before installing
Probe.py re-executes itself under the Hermes Python interpreter located via HERMES_PYTHON / VIRTUAL_ENV / the parent process or the Hermes installer's managed venv (HERMES_HOME/installs). Every credential read is read-only and nothing is refreshed or stored: it reads provider API keys from the process environment, the Hermes .env, the Vaultwarden plugin's local secret cache (cache/vaultwarden_cache.json, only when Vaultwarden is the sole enabled secret source, keyed by a hash of BW_SESSION, which is never sent) and Docker s6 container env files; Hermes's saved OAuth logins (auth.json credential pool and HERMES_HOME/.anthropic_oauth.json); and other coding-assistant CLIs' local logins: Claude Code (credentials file, macOS keychain or Windows Credential Manager), Codex, Cursor (macOS keychain and the IDE's state.vscdb, opened read-only), Kimi, Grok, ZCode and Command Code. For GitHub Copilot (on by default) it tries COPILOT_GITHUB_TOKEN / GH_TOKEN / GITHUB_TOKEN (through Hermes get_secret, else the process environment), then the Copilot editor's apps.json / hosts.json, then runs gh auth token (prompts disabled, 3-second limit), trying each token in turn until GitHub accepts one. It s
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install hermes-resetwatch
Said no? Turn it on later:
hermes plugins enable hermes-resetwatch
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Full setup steps, settings and screenshots are on the plugin's official catalog page from Nous Research.
Memory Shield
Community
Keeps your agent from rewriting or wiping what it remembers about you: by default the agent may add facts to the owner profile (USER.md) wherever it may write and may correct one only when the owner asks in a direct…
Keeps your agent from rewriting or wiping what it remembers about you: by default the agent may add facts to the owner profile (USER.md) wherever it may write and may correct one only when the owner asks in a direct chat, the CLI or a trusted chat, and nobody may delete one; the agent's notes (MEMORY.md) cannot be deleted; group chats are read-only except for the people and owner-only chats you trust, and once trusted_users is set a direct chat with anyone not listed is read-only too; Hermes' unattended background review may not touch the owner profile. Block, ask through Hermes' approval prompt or just log, and undo edits and deletions from snapshots.
Good to know before installing
No network calls, credentials, shell commands or background processes. One pre_tool_call hook reads the turn's platform, chat type, chat id and name, sender id and name and the cron flag (gateway.session_context) and whether the run is Hermes' unattended review (tools.skill_provenance.is_unattended_review), checks memory tool calls, and checks write_file, patch, terminal and execute_code calls only for a write into, or a removal of, USER.md, MEMORY.md or the memory folder (reading a memory file or copying it elsewhere passes in calls up to 16 KB; a longer call that names them is refused); mode approve asks through Hermes' approval gate in the owner's direct chats and the CLI and blocks in shared chats, cron and direct chats with people not in trusted_users. Writes plugin-data/memory-shield/audit.jsonl (the last 500 refused or flagged writes: time, chat, sender and a 160-character excerpt) and snapshots/ (full copies of USER.md and MEMORY.md taken right before an allowed edit or deletion, 20 per file by default). Reads the memory files and holds the memory store's lock files while it reads or writes them (not on Windows); /memory-shield restore overwrites one memory file. /memory-sh
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install memory-shield
Said no? Turn it on later:
hermes plugins enable memory-shield
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Sendblue iMessage / SMS integration. Provides typing indicators, async daemon routing, and Active LLM Tools for SMS dispatch.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install sendblue
Said no? Turn it on later:
hermes plugins enable sendblue
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Venice AI web search + extract — privacy-first, zero data retention, $0.01/request.
How to install
Install it. Hermes asks for any API keys it needs (this one needs name), then asks whether to turn it on.
hermes plugins install web-venice
Said no? Turn it on later:
hermes plugins enable web-venice
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Full setup steps, settings and screenshots are on the plugin's official catalog page from Nous Research.
Account Pool
Community
Run several accounts of one provider in Hermes' credential pool: `hermes account <provider> <action>` adds accounts with email verification, shows live usage per account, keeps the least-used account first (fill_first…
Run several accounts of one provider in Hermes' credential pool: `hermes account <provider> <action>` adds accounts with email verification, shows live usage per account, keeps the least-used account first (fill_first + a 5-minute balance timer over every provider with accounts), parks accounts for hours or on a weekly schedule, shows which bot is on which account and its tokens at API prices, alerts on dead logins, a full pool or a runaway bot, and can start an idle account's 5-hour window with one tiny request (wake). Adapters for anthropic and openai-codex; generic commands for any pool provider. Optional Accounts page in Hermes Desktop. Linux and Windows.
Good to know before installing
Logins stay in Hermes: park holds an entry out of rotation in Hermes' pool (mark_exhausted_and_rotate, reason "parked") and unpark is reset_status; the plugin's park list holds labels, ids and times only; expired tokens are refreshed by Hermes (try_refresh_matching); sign-in, in the terminal and on the Desktop page, runs `hermes auth add`; other profiles' accounts are read with read_credential_pool. It uses each account's access token as Hermes stores it for live usage and the email check (api.anthropic.com/api/oauth/usage and /profile, chatgpt.com/backend-api/wham/usage, with the client headers of Hermes' usage fetcher) and for wake (a command, a page button, or the timer after the opt-in `wake auto-on`): one one-word request per account through Hermes' Anthropic client or the Codex Responses endpoint, and the account's model list. Telegram alerts through `hermes send -t telegram` to the bot's home channel or a chat id kept in account-pool.json; no telemetry. Its session tracker, a post_api_request hook, runs in every profile and writes labels, ids and token counts only. `setup` writes plugins.enabled of every profile, links the plugin into every profile, removes the retired accou
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install account-pool
Said no? Turn it on later:
hermes plugins enable account-pool
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Supermemory semantic long-term memory provider — profile recall, semantic search, explicit memory tools, and per-turn conversation capture. Maintained by Supermemory; the same provider that shipped in-tree.
How to install
Install it. Hermes asks for any API keys it needs (this one needs SUPERMEMORY_API_KEY), then asks whether to turn it on.
hermes plugins install supermemory
Said no? Turn it on later:
hermes plugins enable supermemory
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
VK Messenger gateway with community Long Poll, media attachments, allowlists, and project-lane routing.
How to install
Install it. Hermes asks for any API keys it needs (this one needs VK_GROUP_TOKEN, VK_GROUP_ID), then asks whether to turn it on.
hermes plugins install vk-platform
Said no? Turn it on later:
hermes plugins enable vk-platform
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Defuddle content extraction — local, free, no API key. Shells out to the defuddle CLI for clean article markdown. Extract-only; pair with SearXNG or another search provider.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install web-defuddle
Said no? Turn it on later:
hermes plugins enable web-defuddle
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Full setup steps, settings and screenshots are on the plugin's official catalog page from Nous Research.
Cursor Acp
Community
Use your Cursor subscription models (Claude Opus/Sonnet 5.5, Composer 2.5, GPT-5.6, Grok) inside Hermes Agent via the Cursor CLI's ACP server (`agent acp`).
Use your Cursor subscription models (Claude Opus/Sonnet 5.5, Composer 2.5, GPT-5.6, Grok) inside Hermes Agent via the Cursor CLI's ACP server (`agent acp`). Install with one command (`npx hermes-cursor-acp`) or `hermes plugins install cursor-acp`. The plugin rewrites `.cursor/cli.json` in an empty private workspace every run with a deny-all list (`Read`, `Write`, `Shell`, `WebFetch`, `Mcp`) to prevent Cursor's native CLI tools from accessing host files. It includes a custom extractor that recovers both flat `{"name", "arguments"}` and inline `<tool_call>` blocks that standard regex bridges drop, retries on concurrent `set_model` calls, and reports `usage=None` so Hermes proactive context compression functions correctly.
Good to know before installing
Sends the full conversation, Hermes tool schemas and tool results to Cursor through your locally installed, signed-in Cursor CLI (`agent acp`); Cursor's own tools are fenced by a deny-all `.cursor/cli.json` that the plugin rewrites in a private workspace under `$HERMES_HOME/cache/cursor-acp-workspace` on every start (a differing existing file is backed up to `cli.json.bak`), and Hermes executes every tool call under its own approvals.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install cursor-acp
Said no? Turn it on later:
hermes plugins enable cursor-acp
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Full setup steps, settings and screenshots are on the plugin's official catalog page from Nous Research.
Honcho Memory
Community
Desktop workbench for the Honcho memory behind the focused chat: read the newest conclusions, trace their premises, search messages, ask scoped questions, and add confirmed corrections or source files.
Desktop workbench for the Honcho memory behind the focused chat: read the newest conclusions, trace their premises, search messages, ask scoped questions, and add confirmed corrections or source files. Uses each profile's existing Honcho provider configuration.
Good to know before installing
Talks only to the Honcho server each profile is already configured for, through Hermes's own Honcho client and credentials, and polls it about every 10 seconds for the focused chat while Desktop is open. Ask runs a Honcho reasoning call that may use credits. Add fact/correction and Add to session write to Honcho only after you confirm the target in a dialog; Add to session uploads the text or PDF/JSON/text file you choose. A chat from another local profile is read with that profile's own Honcho configuration, only where this plugin is enabled. It reads the focused chat's title and folder from the profile's session database, read-only. No telemetry.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install hermes-honcho-plugin
Said no? Turn it on later:
hermes plugins enable hermes-honcho-plugin
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Watchline delivery adapter for Hermes Agent. Polls a Watchline pull channel and forwards matched events into the Hermes gateway as inbound messages.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install watchline
Said no? Turn it on later:
hermes plugins enable watchline
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Breaking-news ticker for Hermes Desktop: a scrolling RSS/Atom/JSON-Feed strip above the statusbar with per-source favicons, feed search + discovery, grouping modes, offline cache, and in-app reading. Zero API keys, zero model tokens.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install hermes-newswire
Said no? Turn it on later:
hermes plugins enable hermes-newswire
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Full setup steps, settings and screenshots are on the plugin's official catalog page from Nous Research.
Effortless Hermes
Community
Picks the reasoning effort for every prompt: a fast judge (Jev/TypeSafe, a local OpenAI-compatible model, or any custom API) reads the next message once per turn and rewrites the request's reasoning effort before it…
Picks the reasoning effort for every prompt: a fast judge (Jev/TypeSafe, a local OpenAI-compatible model, or any custom API) reads the next message once per turn and rewrites the request's reasoning effort before it ships — easy questions run cheap, hard jobs get high. Port of effortless for Claude Code.
Good to know before installing
Sends the next message (≤4000 chars) plus a short recent-conversation excerpt (≤1600 chars) to the configured judge endpoint (default api.typesafe.ai for Jev; a local or custom OpenAI-compatible API otherwise), solely to pick the effort; the optional model_map setting (off by default) also switches the turn's model to the one you map to the verdict; keeps recent per-turn token counts in its own plugin state; no telemetry; secrets resolved only through Hermes' secret scope; stdlib-only, no pip dependencies.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install effortless-hermes
Said no? Turn it on later:
hermes plugins enable effortless-hermes
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Full setup steps, settings and screenshots are on the plugin's official catalog page from Nous Research.
Memory Pane
Community
A Hermes Desktop pane for the built-in memory: see what the focused chat remembered, updated or forgot, and edit, delete or undo any entry without leaving the conversation.
A Hermes Desktop pane for the built-in memory: see what the focused chat remembered, updated or forgot, and edit, delete or undo any entry without leaving the conversation. Works alongside the built-in memory; not a memory provider.
Good to know before installing
A tool_execution middleware reads MEMORY.md and USER.md around each memory call and logs the entry-level changes to $HERMES_HOME/plugin-data/hermes-memory-pane/; the pane rewrites memory entries only when the user edits, deletes, adds or undoes one, through Hermes's own memory store; no network, credentials, shell commands or background processes.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install hermes-memory-pane
Said no? Turn it on later:
hermes plugins enable hermes-memory-pane
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Generic JSON-file-based bridge adapter. Bridges self-register via a registry/ manifest (presence = registered, rm = unregistered). Polls inbox/ for incoming messages, writes outgoing to outbox/. External services (imsg, Matrix, Telegram, ...) write/read JSON files in a shared directory.
How to install
Install it. Hermes asks for any API keys it needs (this one needs BRIDGE_DIR), then asks whether to turn it on.
hermes plugins install bridge-adapter
Said no? Turn it on later:
hermes plugins enable bridge-adapter
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Live Colorado highway events from CDOT COtrip. No API key. Default corridor is I-70 Golden to east Denver. Separates incidents happening now from scheduled work and unverified Waze reports.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install colorado-traffic
Said no? Turn it on later:
hermes plugins enable colorado-traffic
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Explicit per-profile readiness and native missing-profile preparation; Python installation remains per profile, never automatic on refresh. Profile model settings, default-on Heavy/Light task routing, fallback controls, runtime limits and Fast inheritance. Reads Hermes profile configuration; explicit saves update selected profiles via the native CLI. Routed jobs launch background Hermes children using the user's configured providers/MCP and store private prompts/results locally. No separate credential store, telemetry, self-updater or approval bypass.
Good to know before installing
After you trust and enable the plugin, routing defaults ON for every installed profile without a saved choice, including profiles that installed an earlier version and never saved one (the update turns it on); existing saved OFF and explicit skill opt-outs remain OFF. On enabled profiles, the model-callable `coding_route` tool starts detached background Hermes agents. They get file, terminal and code-execution tools in any existing folder the model names, run under that profile's credentials and approval policy (single-query mode: dangerous commands are refused unless `approvals.single_query_mode: approve`), and their per-job time limit can be set to none. Prompts and results are stored in `plugin-data/model-router/jobs/`. Saving settings writes model, delegation, auxiliary, fallback and `skills.disabled` keys into the config.yaml of the profiles listed in the preview (backups are taken first) and copies a skill into those profiles' `skills/` folders. The Desktop "prepare profiles" action, after you select profiles and confirm, installs and enables this same catalog pin in up to 32 other local profiles through Hermes's plugin manager, where routing then defaults ON. From install, n
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install model-router
Said no? Turn it on later:
hermes plugins enable model-router
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Full setup steps, settings and screenshots are on the plugin's official catalog page from Nous Research.
Memex8
Community
Self-hosted vector memory (Qdrant + TurboQuant compression) with auto-discovered knowledge realms, semantic search across conversations and ingested documents, and a slumber pipeline for memory consolidation.
Self-hosted vector memory (Qdrant + TurboQuant compression) with auto-discovered knowledge realms, semantic search across conversations and ingested documents, and a slumber pipeline for memory consolidation. v1.1.0 adds optional per-session injection dedup with char-budget gate, decision/evaluation memory-type promotion (slower decay for load-bearing turns), prompt-injection sanitization on read, configurable system-prompt template, export/import memories as JSON, and a /memex8 slash command. New memory behaviours are opt-in; /memex8 export|import is confined to $HERMES_HOME/plugin-data/memex8 unless enable_export_import=true.
Good to know before installing
Auto-stores a truncated copy (first 300 chars) of every user and assistant turn, plus per-message previews of the whole transcript at session end and before context compression, to the memex8 server at MEMEX8_BASE_URL (self-hosted; default http://localhost:8080) using MEMEX8_API_KEY.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs (this one needs MEMEX8_API_KEY), then asks whether to turn it on.
hermes plugins install memex8
Said no? Turn it on later:
hermes plugins enable memex8
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
fmsg platform adapter for Hermes Agent. Real-time inbound over the fmsg-webapi WebSocket (/fmsg/ws), outbound via its JSON REST routes. Authenticates with an fmsg API key (fmsgk_...) exchanged for a short-lived JWT — run the agent as a derived sub-account such as @alice_hermes@example.com. Only http
How to install
Install it. Hermes asks for any API keys it needs (this one needs name), then asks whether to turn it on.
hermes plugins install fmsg-platform
Said no? Turn it on later:
hermes plugins enable fmsg-platform
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Search one index you run and watch one saved query, with index_search, index_health and index_watch.
Good to know before installing
Sends MEILI_API_KEY as a Bearer token to your MEILI_HOST. Over plain http that key travels unencrypted, and a master or admin key is sent as given, so use a search-only key. The tools are on for every surface, including gateways: anyone you allow on your Hermes gateway can search with your key and approve a watch. Turn a surface off with hermes tools. index_watch schedule creates one cron job that runs a model turn every 15 minutes (two or more model requests when it calls the tool) with no daily cap, and it watches only the first 20 hits. The job stays after uninstall until you run hermes cron remove, with -p and the profile name when the profile is not default. Scheduling and unscheduling ask for Hermes approval every time and are refused under cron, yolo, approvals off, unattended sessions and plugins.isolation: host, where the watch only posts a notice that it cannot run, on every run. Answering "always" adds an unused line to config.yaml. The watch relies on private parts of Hermes, so an update can stop it.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs (this one needs MEILI_HOST, MEILI_API_KEY), then asks whether to turn it on.
hermes plugins install index-search
Said no? Turn it on later:
hermes plugins enable index-search
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Full setup steps, settings and screenshots are on the plugin's official catalog page from Nous Research.
Refusal
Community
Detects when a model refuses a request or deflects to an educational alternative, and retries the turn on the next model in your fallback_providers chain.
Detects when a model refuses a request or deflects to an educational alternative, and retries the turn on the next model in your fallback_providers chain; optionally appends your own text to the retry's system prompt after a refusal. Detection is local and rule-based (regex patterns over the response prose); tool-calling responses, code blocks and partial answers are treated as genuine answers.
Good to know before installing
An llm_execution middleware inspects every model response before the agent loop uses it, and when it detects a refusal it sets finish_reason to content_filter (Anthropic: stop_reason to refusal) so Hermes activates its fallback model; this is logged and counted identically to a real provider-signalled refusal (reason=content_policy_blocked), with no way for a plugin to register a distinct reason, so the detection log line carries session/model/provider/api_request_id to let the two be told apart by hand; a triggered fallback resends the full conversation to the fallback provider and spends tokens there; with no fallback provider configured, Hermes replaces the response with its standard safety-refusal notice, so a false positive discards a valid answer. Optional (off by default, empty preprompt/preprompt_escalated settings): an llm_request middleware appends your configured text to the system prompt of requests in an armed session, so a same-route retry of the same model carries it; armed state is memory-only unless persist_state is enabled, in which case session ids/levels/turn ids (never text or credentials) are kept in Hermes's plugin state. No network calls, no files read or wr
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install refusal
Said no? Turn it on later:
hermes plugins enable refusal
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Long-term memory for Hermes, backed by HydraDB. Relevant context is recalled before each turn, and completed turns are saved in the background.
Good to know before installing
When selected as the memory provider, each turn's user message (as the recall query) and every completed user/assistant turn are sent over HTTPS to the HydraDB API (api.hydradb.com, or HYDRADB_BASE_URL) under HYDRADB_API_KEY.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs (this one needs HYDRADB_API_KEY, HYDRADB_DATABASE), then asks whether to turn it on.
hermes plugins install hydradb
Said no? Turn it on later:
hermes plugins enable hydradb
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Connect a Hermes agent through an existing Ando invitation and receive authorized messages.
Good to know before installing
Connects outbound to Ando (mcp.ando.so, api.ando.so, realtime.ando.so) with an agent API key stored at $HERMES_HOME/ando/connection.json (mode 600); Ando messages from the connecting member (or members the operator explicitly allowlists) start Hermes turns and the final reply, with any files it attaches, is posted back to Ando; message attachments are uploaded and downloaded through short-lived single-use links Ando issues (up to 20 MB each; received files are cached in the profile's Hermes cache); hermes ando connect edits gateway.platforms.ando in config.yaml.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install ando-platform
Said no? Turn it on later:
hermes plugins enable ando-platform
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Research search for web_search and web_extract — papers, books, patents, standards and Wikipedia, with citable source links (Machine Library API key).
How to install
Install it. Hermes asks for any API keys it needs (this one needs MACHINELIBRARY_API_KEY), then asks whether to turn it on.
hermes plugins install machine-library
Said no? Turn it on later:
hermes plugins enable machine-library
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Full setup steps, settings and screenshots are on the plugin's official catalog page from Nous Research.
Antigravity Agy
Community
Run Hermes on a Google Antigravity subscription through the official agy CLI, with agy as the model only: a generated custom agent removes agy's built-in tools, Hermes tools are exposed through an inventory-only MCP…
Run Hermes on a Google Antigravity subscription through the official agy CLI, with agy as the model only: a generated custom agent removes agy's built-in tools, Hermes tools are exposed through an inventory-only MCP server with native function calling, and every tool call is handed back to Hermes. Fails closed on any non-Hermes tool step or silent agent fallback; pre-warms agy for fast tool loops.
Good to know before installing
Routes inference through Google's official agy CLI and counts against the Antigravity/Gemini subscription agy is signed in to; the plugin never reads, copies or links agy's credential files. agy runs in your real HOME with the gateway's environment; by default the plugin moves the agy conversation records created by its own idle pre-warmed processes under ~/.gemini/antigravity-cli/ to the trash (trash-put from PATH; ANTIGRAVITY_AGY_CLEANUP=off disables) and pre-warms up to 4 idle agy processes (~400 MB each).
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install antigravity-agy
Said no? Turn it on later:
hermes plugins enable antigravity-agy
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Hosted memory for Hermes with a full history: see what your agent remembered, when it changed and why. Recalls relevant memories before each turn, adds bilinc_recall, bilinc_remember, bilinc_revise and bilinc_forget tools, and mirrors Hermes built-in memory writes. Maintained by ReARC Labs; free plan available.
Good to know before installing
Recall queries (the text of each message, when auto_recall is on), tool arguments, mirrored built-in memory writes, the Hermes session id and, only if memory.bilinc.capture_turns is turned on, each turn's user and assistant text are sent over HTTPS to bilinc.space under your BILINC_API_KEY; requests carry User-Agent bilinc-hermes/<version> and X-Bilinc-Client: hermes-plugin headers. Secret-looking strings are redacted before sending. Recall and writes run on background threads inside Hermes; no separate process, local database or telemetry.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs (this one needs BILINC_API_KEY), then asks whether to turn it on.
hermes plugins install bilinc
Said no? Turn it on later:
hermes plugins enable bilinc
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Carbon Voice gateway adapter for Hermes Agent. Connects via Socket.IO (primary) with REST polling fallback — no public webhook or tunnel required. Persists a cursor to disk so messages received while Hermes is offline are processed on the next startup. Replies via POST /v3/messages/start, so the Her
How to install
Install it. Hermes asks for any API keys it needs (this one needs name), then asks whether to turn it on.
hermes plugins install carbonvoice-platform
Said no? Turn it on later:
hermes plugins enable carbonvoice-platform
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
MiniMax web search via the Token Plan API — keyed, no CAPTCHA
How to install
Install it. Hermes asks for any API keys it needs (this one needs MINIMAX_API_KEY), then asks whether to turn it on.
hermes plugins install minimax-search
Said no? Turn it on later:
hermes plugins enable minimax-search
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Command Code CLI-key model provider with a live OAuth catalog property bridge, plan limits and failover classification.
Good to know before installing
Uses the CLI-internal api.commandcode.ai/alpha/generate endpoint and CLI client headers, sends the full conversation, tool schemas, cwd path and OS/Python version, and imports a static key from the read-only CLI credential file or a browser hand-off; no OAuth exchange takes place. Hermes also bundles commandcode for the documented Provider API.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install commandcode-oauth
Said no? Turn it on later:
hermes plugins enable commandcode-oauth
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Full setup steps, settings and screenshots are on the plugin's official catalog page from Nous Research.
Ctx checkpoint
Community
Memory provider that archives the direct user/assistant transcript into a ctx knowledge store before every compaction, deterministically and with no LLM in the path, and hands the resulting block IDs to the compaction…
Memory provider that archives the direct user/assistant transcript into a ctx knowledge store before every compaction, deterministically and with no LLM in the path, and hands the resulting block IDs to the compaction summary so the exact wording stays recoverable. Best-effort on any supported Hermes; with compression.checkpoint_required on Hermes 0.21.1+ it uses the fail-closed pre-compress checkpoint contract (#94639) and blocks compaction until the checkpoint is durable.
Good to know before installing
The turns about to be compacted (tool results and system prompts excluded, secrets redacted twice) are written through Hermes' MCP dispatch to the ctx server you configure in mcp_servers; the plugin holds no credentials, makes no other network calls and has no telemetry; it imports the core modules agent.redact, tools.registry and hermes_cli.config.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install ctx_checkpoint
Said no? Turn it on later:
hermes plugins enable ctx_checkpoint
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Meta web search — routes web_search through the server-side web_search tool on Meta's Responses API (Muse Spark); search-only.
Good to know before installing
Each search is a separate billed Meta Model API call (per-query web search fee plus tokens) on MODEL_API_KEY.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs (this one needs MODEL_API_KEY), then asks whether to turn it on.
hermes plugins install web-meta-ai
Said no? Turn it on later:
hermes plugins enable web-meta-ai
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Answers an allowlisted direct message on your Meshtastic® radio.
Good to know before installing
Talks only to the MESHTASTIC_URL you set, over plain tcp, so channel keys and text travel unencrypted. An empty MESHTASTIC_ALLOWED_NODES answers nobody. A channel packet or a direct message without pkiEncrypted is never answered, because a channel key can forge a sender. Allowlisted nodes can prompt the agent, use tools until hermes tools turns them off, and approve one command at a time. An approval question is sent only when its line shows exactly the command /approve will run. A line that still goes out after that send was declined does not let /approve run it, and the next approval is declined once. Each direct message runs an agent turn on your model, and the agent cannot send on its own. No cron job comes with it; one the agent makes stays until hermes cron remove. Hermes main does not load this platform when plugins run in their own process. Node ids, sizes and times go to nodes.json in this profile's plugin-data and stay after uninstall. The message text is not stored. Not tried on a physical radio. Meshtastic® is a registered trademark of Meshtastic LLC. This site is not affiliated with or endorsed by the Meshtastic project.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs (this one needs MESHTASTIC_URL), then asks whether to turn it on.
hermes plugins install radio-dm-gateway
Said no? Turn it on later:
hermes plugins enable radio-dm-gateway
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Experimental model transport using the official Cursor Agent SDK. Requests go to Cursor through its official SDK. The plugin cannot enforce an included-only spending cap and never changes account spending settings. Instructions use a Cursor project rule, not a native system-role override. Permission and prompt-hierarchy equivalence remain unqualified. The plugin spawns Node and a local SDK agent, runs Git initialization in a plugin-owned workspace, and writes project rules, hooks, and SDK state under its own runtime directory. That state may contain private conversation data and currently remains on disk. Credentials come from Hermes's profile-scoped secret accessor. No credential store or vendor token file is read directly by plugin code. Text-only, buffered responses. The bundled model snapshot is not live entitlement discovery. Optional picker hooks require Hermes support. This plugin does not patch the desktop.
Good to know before installing
Spawns Node and a local Cursor Agent SDK worker per request (`npm ci` in the plugin dir installs Cursor's official native platform binaries from npm); sends the full conversation, system prompt and tool schemas to Cursor with your CURSOR_API_KEY; Cursor-native shell/file tools are denied by a fail-closed project hook and the run is cancelled if one fires, but that enforcement is not proven in every Cursor runtime; each request's workspace and conversation state stay under the plugin's runtime/ directory until you delete them.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs (this one needs CURSOR_API_KEY), then asks whether to turn it on.
hermes plugins install cursor-sdk
Said no? Turn it on later:
hermes plugins enable cursor-sdk
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Recall and save decisions, preferences and facts in your Kumiho Cloud workspace. Portable Agent Plugins v1 package with hosted OAuth MCP and six skills.
Good to know before installing
Connects to Kumiho's hosted MCP at https://mcp.kumiho.cloud/mcp (OAuth at control.kumiho.cloud, tokens held by Hermes); memory queries, including an automatic topic lookup near the start of relevant conversations, and approved captures are sent there; the server is marked `trust: untrusted`, so Hermes asks before every tool not annotated read-only. The optional backfill helper reads only files the user names and writes review packets to a new folder; Dream State can create a recurring Hermes cron task when asked (read-only by default). No local server, telemetry or background process.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install kumiho-memory
Said no? Turn it on later:
hermes plugins enable kumiho-memory
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Serply web search provider for Hermes Agent. Google web results via the Serply SERP API. Requires SERPLY_API_KEY (sign up at https://serply.io).
How to install
Install it. Hermes asks for any API keys it needs (this one needs SERPLY_API_KEY), then asks whether to turn it on.
hermes plugins install web-serply
Said no? Turn it on later:
hermes plugins enable web-serply
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Full setup steps, settings and screenshots are on the plugin's official catalog page from Nous Research.
Reaction Feedback
Community
Lets the agent see the emoji reactions you leave on its Telegram messages: react 👍 or 👎 to a reply and the agent's next turn in that private chat starts with a short note naming the reaction and the message it followed.
Lets the agent see the emoji reactions you leave on its Telegram messages: react 👍 or 👎 to a reply and the agent's next turn in that private chat starts with a short note naming the reaction and the message it followed.
Good to know before installing
Registers pre_gateway_dispatch (always returns None, so dispatch is never changed), pre_llm_call and gateway_platform_event hooks; keeps a bounded in-memory record of private-chat message ids, 60-character excerpts and reactions; writes nothing to disk, makes no network requests, starts no processes and reads no secrets.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install reaction-feedback
Said no? Turn it on later:
hermes plugins enable reaction-feedback
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Full setup steps, settings and screenshots are on the plugin's official catalog page from Nous Research.
OpenPaths
Community
OpenPaths model provider for Hermes: OpenAI-compatible inference at openpaths.io with auto, auto-easy-task, auto-medium-task, auto-hard-task and auto-think task-tier routes plus GPT, Claude, Gemini and DeepSeek models…
OpenPaths model provider for Hermes: OpenAI-compatible inference at openpaths.io with auto, auto-easy-task, auto-medium-task, auto-hard-task and auto-think task-tier routes plus GPT, Claude, Gemini and DeepSeek models, live /models listing, and Hermes reasoning effort mapped to OpenPaths reasoning_effort.
Good to know before installing
Prompts, tool schemas, tool results and conversation context are sent to openpaths.io with your OPENPATHS_API_KEY and forwarded to the upstream model vendors it routes to; the plugin makes no other network calls and registers no tools or hooks.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs (this one needs OPENPATHS_API_KEY), then asks whether to turn it on.
hermes plugins install openpaths
Said no? Turn it on later:
hermes plugins enable openpaths
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Review pending Hermes memory writes and inspect, add, edit, delete, or AI-compact stored memory from Desktop.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install magi
Said no? Turn it on later:
hermes plugins enable magi
Restart Hermes so it loads the plugin.
Some plugins also need a sign-in or settings: see Full setup & settings. Memory plugins ask whether to become your memory provider instead. Steps follow the official plugins guide↗ (opens in a new tab).
Full setup steps, settings and screenshots are on the plugin's official catalog page from Nous Research.
No plugins match that search.
Read before you install
Plugins run code inside your agent. Official plugins come from Nous Research's own repo. Community plugins are independent projects, so check the source, the permissions they ask for and any API keys they need. New to plugins? Start with skills vs plugins vs MCP.
