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.
117 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.
Mcp Serve
Official
Run Hermes as an MCP server: `hermes mcp-serve` lets any MCP client (Claude Code, Claude Desktop, Cursor, Codex) list and read your messaging conversations, poll live events, send messages and answer approvals over stdio.
Run Hermes as an MCP server: `hermes mcp-serve` lets any MCP client (Claude Code, Claude Desktop, Cursor, Codex) list and read your messaging conversations, poll live events, send messages and answer approvals over stdio. `--tools` (off by default) also exposes Hermes' web, browser, vision, image generation, skills and TTS tools; terminal, file and kanban tools are never exposed. Formerly built into Hermes core as `hermes mcp serve` (MCP clients change their args to ["mcp-serve"]).
Good to know before installing
The connecting MCP client reads the active profile's conversation history (state.db) and can send messages as your connected platform accounts; with `--tools` it also runs Hermes tools under your configured provider credentials, which can spend provider credits. The plugin itself makes no network or model calls beyond those tool calls and sends.
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 mcp-serve
Said no? Turn it on later:
hermes plugins enable mcp-serve
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).
Call any Scalekit connector from Hermes: 500+ connectors and 20,000+ tools.
Good to know before installing
HTTP calls go to the Scalekit environment URL you paste; the three Scalekit values are read with get_secret; one person id is stored at plugin-data/scalekit/identifier; the model can request an approval link, run a named tool, or proxy one HTTP request. Provider tokens stay in Scalekit; no shell, no background process, no telemetry.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs (this one needs SCALEKIT_CLIENT_ID, SCALEKIT_CLIENT_SECRET, SCALEKIT_ENVIRONMENT_URL), then asks whether to turn it on.
hermes plugins install scalekit
Said no? Turn it on later:
hermes plugins enable scalekit
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.
Bubblewrap
Official
Bubblewrap terminal backend for Linux: every terminal and execute_code command runs in its own bwrap sandbox on this machine at the host's own paths, with a read-only root, a writable working directory (per profile…
Bubblewrap terminal backend for Linux: every terminal and execute_code command runs in its own bwrap sandbox on this machine at the host's own paths, with a read-only root, a writable working directory (per profile: restricted, workspace or network), the dot entries of the home directory (credential stores always) and the Hermes home hidden, private temp and process directories, agent and docker sockets masked, and per-command memory, CPU and process limits (`hermes config set terminal.backend bubblewrap`).
Good to know before installing
No Python dependencies and no network calls of its own: it runs your commands through the `bwrap` and `prlimit` binaries you install (bubblewrap, util-linux), and under the default `network` profile those commands share the host network. It reads no credentials. File tools (read_file, write_file, patch, search_files) still run on the host and are not confined, and because the working directory is a real host directory Hermes keeps its dangerous-command prompts, as for the local backend.
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 bubblewrap
Said no? Turn it on later:
hermes plugins enable bubblewrap
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).
Claude Code-style plan mode for Hermes: the agent explores and writes a plan, and edits nothing else until you approve it. It works on every interactive Hermes surface: /plan or /planmode on blocks mutating Hermes tools dispatched through pre_tool_call (writes outside .hermes/plans, terminal, code execution, delegation, MCP) while the agent explores, asks clarifying questions and writes a plan. The agent shows the complete plan, submits it with plan_mode(submit) and one approval card asks: approve and auto-accept edits, approve and review each file edit (when Hermes approvals are on, every write_file and patch then gets Hermes' approval prompt with its path and a preview), or keep planning; Other sends feedback, and the revised plan says what changed since the last revision. Once approved, it implements in the same turn and tracks todo_list progress. To change the plan yourself, keep planning, run /planmode edit, save, then /planmode show and /planmode approve; the text you were shown is saved as an approved snapshot, so later edits to the file do not change the approval. The classic profile (plan_profile: classic), approval_card: false and surfaces without a card use Hermes' own approval prompt (gateway buttons, the TUI/Desktop card, the CLI panel or /approve). When yolo or approvals.mode: off would auto-approve, it asks through clarify instead, so a person still decides. By default plans are compact and decision-complete, the lean profile asks at most three targeted questions, and the agent does not commit unless asked (plan_style, plan_profile, allow_commits and plan_skill settings). Classic CLI enforced on every supported Hermes; gateway, TUI and Desktop from Hermes v2026.9.24 (0.21.5), and refused on older builds; 0.21.6 recommended. On the interactive classic CLI, Hermes' inline diff shows a new plan file in full, so the agent does not retype it (write_once). A report-a-problem skill drafts a bug report and files it with gh only after you approve the draft.
Good to know before installing
Plan-mode does not enforce Codex app-server exec/applyPatch or TUI /background and btw side agents. A plan is approved only on its approval card (a clarify question only the plugin issues and checks), through Hermes' own approval prompt, or with /planmode approve. It reads nine groups of internal Hermes seams for session identity, workspace, profile, skills.inline_shell and the classic CLI's inline diff, each guarded with a documented fallback. It creates .hermes/plans under the session working directory, reads plan files there, saves the approved text of an edited plan as a new file under .hermes/plans/approved, and keeps per-session state in plugin state. It adds a turn note while planning or executing, a ≤200-char system-prompt hint, and a one-line footer to short replies on chat platforms. When a turn ends while a submitted plan still awaits approval and no approval card was answered, it adds one line to that reply on any surface, whatever its length, saying how to approve (unless footer is off). When you pick "Always" on a plan approval or a per-edit review, Hermes core writes a plugin_rule:plan-mode entry to command_allowlist in config.yaml; each entry is keyed to one plan re
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 plan-mode
Said no? Turn it on later:
hermes plugins enable plan-mode
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.
Apify
Community
Apify, the largest marketplace of tools for AI: discover, run, and collect results from thousands of ready-to-run Actors, from the comfort of your agent, plus an Apify backend for the built-in web_search and…
Apify, the largest marketplace of tools for AI: discover, run, and collect results from thousands of ready-to-run Actors, from the comfort of your agent, plus an Apify backend for the built-in web_search and web_extract tools.
How to install
Install it. Hermes asks for any API keys it needs (this one needs APIFY_API_TOKEN), then asks whether to turn it on.
hermes plugins install apify
Said no? Turn it on later:
hermes plugins enable apify
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.
Singularity
Official
Singularity/Apptainer terminal backend: Hermes's terminal, file and code tools run inside a hardened, rootless container on this machine (--containall --no-home), with CPU/memory limits and an optional per-session…
Singularity/Apptainer terminal backend: Hermes's terminal, file and code tools run inside a hardened, rootless container on this machine (--containall --no-home), with CPU/memory limits and an optional per-session persistent overlay (`hermes config set terminal.backend singularity`; image from `terminal.singularity_image`). Formerly built into Hermes core; existing users are migrated automatically.
Good to know before installing
No Python dependencies and no network calls of its own: it drives the `apptainer` (or `singularity`) CLI you install, which pulls the configured image (default `docker://nousresearch/hermes-sandbox:desktop`) and caches a SIF under your scratch directory. Your skills and registered credential files are bind-mounted read-only into the container, and because the container is a disposable sandbox Hermes skips its dangerous-command prompts there, as for its other container backends.
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 singularity
Said no? Turn it on later:
hermes plugins enable singularity
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).
JetBrains Junie coding agent as a Hermes model provider over ACP.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install junie-acp
Said no? Turn it on later:
hermes plugins enable junie-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.
Glasser
Community
Glasser's hosted MCP: search, inspect and run 1,000+ paid data endpoints (people and company enrichment, SEO, web, social media, US property, stock quotes) with one key, pay per call.
Glasser's hosted MCP: search, inspect and run 1,000+ paid data endpoints (people and company enrichment, SEO, web, social media, US property, stock quotes) with one key, pay per call. Portable Agent Plugins v1 package: a stdio bridge through mcp-remote (browser sign-in, no key in the package) plus a bundled skill for setup and the spending rules. Requires Node.js 18+.
Good to know before installing
Paid data broker: each run tool call charges your Glasser balance per the endpoint's published price with no in-Hermes confirm step; runs npx mcp-remote@0.14.3 as a local stdio bridge to https://api.glasser.ai/mcp; the sign-in token is cached by mcp-remote under ~/.mcp-auth/.
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 glasser
Said no? Turn it on later:
hermes plugins enable glasser
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).
Browse Tailscale devices and connect to remote Hermes instances in Hermes Desktop.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install hermes-tailscale
Said no? Turn it on later:
hermes plugins enable hermes-tailscale
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.
Webly
Community
Publish a website, or share files and folders, at a live *.webly.site URL from Hermes Agent with no account, then sign in to claim it, edit it, and publish drafts the person approves through the hosted Webly MCP server…
Publish a website, or share files and folders, at a live *.webly.site URL from Hermes Agent with no account, then sign in to claim it, edit it, and publish drafts the person approves through the hosted Webly MCP server (Streamable HTTP). Portable Agent Plugins v1 package (plugin.json + mcp.json + skills/).
Good to know before installing
MCP tool calls go to api.webly.ai/v1/mcp, sending WEBLY_API_KEY as a bearer header when it is set, and the webly server is marked untrusted so Hermes asks before each Webly MCP tool that changes something. The bundled webly skill runs a local Node.js 20+ helper (scripts/webly.mjs) through the shell: it uploads only the files and folders the person asks to publish or share to api.webly.ai (sites are public unless password-protected), runs the project's npm/pnpm/yarn install and build first when the folder has a build script, opens the browser for sign-in and claim, and keeps its anonymous-site token and optional OAuth sign-in in ~/.webly (mode 0600). Its doctor command reads ~/.claude.json, ~/.claude/settings.json and ~/.codex/config.toml only to see whether a webly MCP server is configured there. Publishing, deleting and replacing sites through that helper are gated by skill instructions, not by Hermes approvals. The catalog copy has no update check.
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 webly
Said no? Turn it on later:
hermes plugins enable webly
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.
Looot
Community
looot gives Hermes one sign-in and one prepaid balance for 2,500+ data API endpoints from 90+ providers: work emails and phones, company and people search, Google results, web pages as markdown, news, local businesses.
looot gives Hermes one sign-in and one prepaid balance for 2,500+ data API endpoints from 90+ providers: work emails and phones, company and people search, Google results, web pages as markdown, news, local businesses. Portable Agent Plugins v1 package with the hosted MCP server (remote streamable-http, OAuth 2.1 with dynamic client registration) and 21 skills.
Good to know before installing
The run tool spends your prepaid looot balance (Hermes asks before each paid call and refuses them in unattended runs); the names, domains and emails you look up are sent to looot.ai and its upstream data providers; a `looot` 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 looot
Said no? Turn it on later:
hermes plugins enable looot
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).
Read-only Unraid monitoring for Hermes Desktop: server metadata, array capacity, Docker containers, virtual machines, and grouped alerts.
Good to know before installing
Sends fixed GraphQL queries to the user-configured Unraid host and stores a VIEWER key through Hermes secret scope; no telemetry, background process, shell, SSH, MCP, or mutations.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs (this one needs UNRAID_API_KEY), then asks whether to turn it on.
hermes plugins install unraid
Said no? Turn it on later:
hermes plugins enable unraid
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 Tools
Community
Connect AgentPlaybooks skills and hosted MCP/OpenAPI tools to Hermes, with guided setup, profile selection, permission diagnostics and an authenticated MCP check.
Connect AgentPlaybooks skills and hosted MCP/OpenAPI tools to Hermes, with guided setup, profile selection, permission diagnostics and an authenticated MCP check. Requires Node.js 20+ and @agentplaybooks/cli@0.7.0; the CLI is installed separately.
Good to know before installing
Skill-only; the separately installed apb CLI does the work. `apb hermes setup --apply` restores complete skill trees into the chosen project and edits the selected Hermes profile's config.yaml to register those skills and a hosted playbook MCP endpoint. Local writes require --apply; setup also probes MCP initialization and tool listing using a scoped key from the process environment, without invoking tools. `doctor` checks profile credential variable names without exposing their values. Hosted operations contact agentplaybooks.ai (or an explicitly configured service); skill restore uses `apb login`, and `apb push --apply` uploads project configuration (secret references, not values).
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 agentplaybooks-tools
Said no? Turn it on later:
hermes plugins enable agentplaybooks-tools
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 28 production-ready MCP servers to any coding agent. Portable Agent Plugins v1 package (mcp.json + skills/): three zero-config stdio servers out of the box (Hacker News, arXiv, npm registry, each pinned to an exact npm version), plus the umt-mcp skill that drives the umt CLI to wire any of the 28 servers into Claude Code, Cursor, Codex, Windsurf, Zed and more. Needs Node.js and npm on PATH.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install umt
Said no? Turn it on later:
hermes plugins enable umt
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).
Frontend design guidance with sixteen original style grammars, twelve craft references, dated modern-web capability recipes, explicit direction context, and stateless tools for bounded source inspection, color contrast, and navigation task paths.
Good to know before installing
The hardlight_inspect tool reads frontend source files (.html, .css, .js, .ts, .jsx, .tsx, .vue, .svelte and similar; at most 500 files and 8 MiB) under any local directory the model names, skipping hidden, symlinked, vendor/build and secret-named paths, and returns their design tokens and findings to the model; the plugin adds a short system-prompt section naming its tools to every session; the tools make no network calls, run no subprocesses and write no files; the optional hardlight browser command, run by hand from a shell, launches a locally installed Chromium against loopback or file:// pages only and writes its report to the output folder you give it.
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 hardlight-ui
Said no? Turn it on later:
hermes plugins enable hardlight-ui
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).
Manage SSH connections and remote workspaces in Hermes Desktop.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install hermes-ssh
Said no? Turn it on later:
hermes plugins enable hermes-ssh
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.
Monid
Community
Connect your agent to third-party tools and APIs (web search and scraping, people and company data, social media, media generation, email and phone) through one integration.
Connect your agent to third-party tools and APIs (web search and scraping, people and company data, social media, media generation, email and phone) through one integration. The agent discovers, compares and runs tools at runtime, paid per call from one Monid balance.
Good to know before installing
Monid (monid.ai, by Monid / monid-ai) is a paid third-party marketplace: the agent installs the @monid-ai/cli npm package or connects to mcp.monid.ai, sends your requests and any uploaded files to Monid and on to its providers, and each run spends your prepaid Monid workspace balance. The "confirm the price first" step is an instruction to the model in the bundled skill, not a Hermes approval prompt, unless you add `trust: untrusted` to `mcp_servers.monid`.
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 monid
Said no? Turn it on later:
hermes plugins enable monid
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).
Deploy, debug and manage Render applications with 20 unchanged skills from render-oss/skills and Render's hosted MCP connection.
Good to know before installing
Connects Hermes to Render's hosted MCP server (mcp.render.com) with your RENDER_API_KEY; the server is marked untrusted so tools without a read-only annotation need approval; account permissions apply and changes can incur Render charges; bundled skills guide shell commands, file edits and Render CLI installs (including the vendor's curl | sh installer).
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs (this one needs RENDER_API_KEY), then asks whether to turn it on.
hermes plugins install render-platform
Said no? Turn it on later:
hermes plugins enable render-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).
Build, run and test Python in an offline container per chat, with results as a plain file:line issue list. The agent syncs files or folders in as copies, then runs run (a shell command), check (ruff error rules) and test (pytest); fixes stay in the box until the user asks for export, which writes changed files back through write_file and reports host files changed since the sync as conflicts. In Hermes Desktop a VS Code-style editor opens beside the chat: file tabs with in-place editing, Run / Check / Test / Stop, and PROBLEMS / OUTPUT / RUNS panels that follow the agent live. Ships the runbox:dev-loop starting-point skill.
Good to know before installing
Needs Docker (on macOS a dedicated colima profile 'runbox', started on demand with --activate=false and its own Docker config); pulls the container image pinned by digest and installs a hash-pinned checker toolchain (ruff 0.16.9, pytest 9.1.1 and deps) from PyPI once per image; the work container has no network, hardened flags, and runs commands without Hermes' dangerous-command approval like Hermes' Docker backend; install (PyPI) always asks for approval in a separate installer container; a write_file/patch on an imported original asks for approval (guard_imports); export reads each existing target through read_file before replacing it; a background thread removes idle containers (15 min), stops the VM, and removes boxes of deleted or long-idle chats (reads state.db read-only hourly); adds a short per-turn note (about 80 tokens); no credentials or 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 runbox
Said no? Turn it on later:
hermes plugins enable runbox
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).
Delegate coding tasks to the Devin CLI, on your computer or in Devin Cloud, with live progress cards and change review in Hermes Desktop. Every launch goes through Hermes approval; Cloud runs need a separate Cloud-managed-permissions confirmation.
Good to know before installing
Spawns the vendor devin CLI from PATH or cli_path (devin acp, devin acp --cloud, devin --print, devin auth status, devin models list); the CLI edits files and runs shell commands in a folder the model picks (any folder unless allowed_roots is set) under the configured permission mode (accept-edits/smart auto-allow edits, dangerous auto-allows everything and needs allow_dangerous: true), and talks to Cognition's servers; Devin Cloud runs use Cognition-managed permissions that Hermes does not gate. Every model-initiated launch goes through Hermes approval; follow-ups typed on a Desktop run card and answers to Devin's prompts on the card start or allow actions on that click without a Hermes prompt (unanswered card prompts are denied after 300 s by default). Uses the existing Devin CLI login and stores no credentials; Hermes provider, bot and tool secrets are stripped from the child environment (extra_env is passed as set). Background runs continue after the turn up to their timeout and post Devin's output back into the conversation as a message when they finish. Run history (prompts, diffs, command output) and model preferences are saved under plugin-data/devin unless persist_runs is
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 devin
Said no? Turn it on later:
hermes plugins enable devin
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).
Run Hermes terminal, file and execute_code tools in Cloudflare Containers, with the workspace saved to a snapshot when idle.
Good to know before installing
You deploy the bundled Cloudflare Worker to your own account first; running containers is billed by Cloudflare; commands and file contents run in the container at HERMES_CF_ENDPOINT using HERMES_CF_TOKEN; the container has outbound internet by default (ENABLE_INTERNET); host env vars, skills and credentials are not copied into the container; dangerous-command approval stays on.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs (this one needs HERMES_CF_ENDPOINT, HERMES_CF_TOKEN), then asks whether to turn it on.
hermes plugins install cloudflare-sandbox
Said no? Turn it on later:
hermes plugins enable cloudflare-sandbox
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 Cron
Community
Operating skill for Hermes cron: lifecycle rules (pause vs disable vs remove, `in 30m` vs `30m`), four ways to avoid needless LLM runs (no_agent, wakeAgent pre-check, monitor, [SILENT]), verify-after-change, a layered…
Operating skill for Hermes cron: lifecycle rules (pause vs disable vs remove, `in 30m` vs `30m`), four ways to avoid needless LLM runs (no_agent, wakeAgent pre-check, monitor, [SILENT]), verify-after-change, a layered runbook and a read-only doctor that complements `hermes cron doctor` and reports one error across several jobs as one scheduler fault and a cron store that cannot be written. Portable Agent Plugins v1 package (plugin.json + skills/); no tools, hooks, middleware or env vars.
Good to know before installing
Skill only, no network; the bundled doctor reads ~/.hermes/cron/jobs.json, ticker stamps and executions.db (or a named profile's store with --profile) read-only, checks the cron directory's permissions and free space without writing, and runs the local hermes cron doctor (argument list, no shell), whose job loader may repair malformed jobs.json entries (--no-builtin skips it).
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-cron
Said no? Turn it on later:
hermes plugins enable hermes-cron
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 Windows Ops
Community
Windows and WSL triage skills for Hermes: processes that survive a quit (orphaned stdio MCP trees, lingering python/uv/node children), WSL-to-Windows interop where an argument reaches cmd.exe as shell syntax, config…
Windows and WSL triage skills for Hermes: processes that survive a quit (orphaned stdio MCP trees, lingering python/uv/node children), WSL-to-Windows interop where an argument reaches cmd.exe as shell syntax, config and encoding traps (UTF-8 BOM, legacy console code pages), and long-path or reserved-name errors. One skill plus a copy-paste command reference; the process, argv and job-object failure modes are pinned to reports #131561, #132069 and #61059 and re-verified against current main.
Good to know before installing
Instructions only: no tools, hooks, middleware, network calls or credentials, and the destructive diagnostics are marked and gated on asking the user first.
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-windows-ops
Said no? Turn it on later:
hermes plugins enable hermes-windows-ops
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).
CLI-backed skills for local code discovery and entity-diff review with Sem, plus optional Weave branch-merge previews. Requires user-installed CLIs; does not apply edits or merges, install binaries, or change Git configuration.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install sem-weave
Said no? Turn it on later:
hermes plugins enable sem-weave
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 Wsl Terminal
Community
Run Hermes' terminal tools inside an existing WSL distro on Windows: one bash -c per command as the distro's default user, over the same files Windows sees at /mnt/<drive>, with no Docker Desktop, no image and no…
Run Hermes' terminal tools inside an existing WSL distro on Windows: one bash -c per command as the distro's default user, over the same files Windows sees at /mnt/<drive>, with no Docker Desktop, no image and no container to start or tear down. A host cwd is translated (C:\Users\me\proj becomes /mnt/c/Users/me/proj, \\wsl.localhost paths become the distro's own paths) and the mount root is asked once with wslpath, so a distro that moved its automount point still gets correct paths. cwd persists across commands, stdin and exit codes ride through wsl.exe, and hermes doctor / setup and the dashboard picker get real rows (wsl.exe path, WSL version, distros present, selected distro). HERMES_WSL_DISTRO and HERMES_WSL_USER select the distro and the user.
Good to know before installing
This is a full-access backend, not a sandbox: when it is the selected backend, everything the terminal tool runs executes as your user inside your distro, with that distro's files, permissions and network access, so dangerous-command approval stays on and Hermes never stops the distro. The plugin itself reads no credentials, stores nothing and makes no network calls; it spawns wsl.exe per command and wslpath once per distro. It rewrites one core-produced string, the trailing cwd marker, because under WSL a command substitution is handed the directory wsl.exe started in rather than the shell's current one (measurements in the repo's VALIDATION.md); nothing in Hermes is patched or rebound at runtime.
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-wsl-terminal
Said no? Turn it on later:
hermes plugins enable hermes-wsl-terminal
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.
Git Review
Community
Git review pane for Hermes Desktop: every changed file of the agent's last turn, uncommitted, staged or unstaged work, the branch since it left the default branch, or any selection of commits, in one scroll, unified or…
Git review pane for Hermes Desktop: every changed file of the agent's last turn, uncommitted, staged or unstaged work, the branch since it left the default branch, or any selection of commits, in one scroll, unified or split, with syntax colours; cites changed lines into the composer as @line: references; and stages, unstages, discards, commits, pushes and opens pull requests from the pane.
Good to know before installing
Reads run git subprocesses (rev-parse, status, diff, log, show, cat-file, ls-files, for-each-ref, branch, remote, rev-list, merge-base, symbolic-ref, hash-object) with timeouts and output caps, external diff and textconv drivers disabled. Writes run only when the user clicks: git add, restore, reset, rm, commit and push, on paths validated inside the repository; discard asks first in Desktop's confirmation dialog and deletes only the untracked files it names. By default hooks are skipped and any write that would run a program named by the repository's own config (filter drivers, signing, SSH or credential commands) is refused; the run_repo_programs setting allows them, and the read_only setting hides and refuses every write. Network use: push to the branch's own name on its remote, and the user's own gh or glab CLI for the PR/MR badge and for Create pull request; the plugin itself makes no network requests. Generate commit message sends what the commit will contain and recent commit subjects to the focused chat's model through Hermes's commit_message one-shot template, only when clicked. It reads files and history inside authorized repositories only (the focused workspace, configur
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 git-review
Said no? Turn it on later:
hermes plugins enable git-review
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 Wslc Sandbox
Community
Run Hermes' terminal, file and execute_code tools in a Linux container on Microsoft's WSL Containers (wslc) runtime, so Windows users who run WSL Containers get a sandbox without Docker Desktop.
Run Hermes' terminal, file and execute_code tools in a Linux container on Microsoft's WSL Containers (wslc) runtime, so Windows users who run WSL Containers get a sandbox without Docker Desktop.
Good to know before installing
The plugin spawns the wslc CLI (one short-lived process per command, plus one long-lived container per environment) and every call runs inside that container; containers have outbound network access unless docker_network is false; persistent sandboxes are left stopped on disk until removed; host environment variables enter the container only when named in docker_forward_env, and only paths listed in docker_volumes are bind-mounted. wslc rejects --cap-drop, --security-opt, --pids-limit, --init and --storage-opt, so this sandbox is not capability-dropped the way the Docker backend's is, container_disk is reported but not enforced, and dangerous-command approval deliberately stays enabled for this backend. The plugin reads no credentials, stores none, and makes no network calls of its own.
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-wslc-sandbox
Said no? Turn it on later:
hermes plugins enable hermes-wslc-sandbox
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).
YAGNI coding and code-review guidance for Hermes: reduce over-engineering, reuse existing behavior, and preserve requirements. One static prompt section; no tools or hooks.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install hermes-minimizer
Said no? Turn it on later:
hermes plugins enable hermes-minimizer
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).
Community plugin that rewrites Hermes terminal commands through RTK (rtk-ai/rtk, Apache-2.0) so the model sees compact output. Adds a backend gate: it rewrites only for terminal backends a profile opts in to (default local), so a Docker or remote backend without rtk never receives an rtk command. Fail-open pre_tool_call hook (2s timeout).
Good to know before installing
Every terminal command on an opted-in backend (default local) is passed to the rtk binary on the Hermes host's PATH (or RTK_BIN) and replaced with rtk's stdout before approval and execution; the plugin does not validate the rewrite, fails open on timeout or error, and downloads nothing (the repo's optional Containerfile fetches a checksum-pinned RTK release only when you build it). Not affiliated with rtk-ai.
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 rtk-rewrite-docker
Said no? Turn it on later:
hermes plugins enable rtk-rewrite-docker
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.
Lightkeeper
Community
An Obsidian-compatible map of how your codebase works (modules, flows, integrations, platform traps, decisions), kept outside the repo and kept honest by hooks: a primer on the first turn, pointers to the notes that…
An Obsidian-compatible map of how your codebase works (modules, flows, integrations, platform traps, decisions), kept outside the repo and kept honest by hooks: a primer on the first turn, pointers to the notes that own what you are asking about or editing, and a reminder through the verify gate when code moved but the map did not. Eight lightkeeper:* skills carry the procedures (init, seed, tidy, trace, release, knowledge-vault, vault-sync, vault-tidy); the vault CLI (find, check, sync, status, log, export) works from any shell. Not a memory provider.
Good to know before installing
Each hook runs python3 on the plugin's own scripts as a synchronous subprocess with a 10 s timeout (on every user message, after every write_file and patch, at the verify step and at session end), and those scripts run read-only git commands (rev-parse, remote get-url, worktree list, ls-tree, status, log) in the session's checkout and in the vault folder; the session primer and prompt hints put text from the vault's notes into the model's context; the hooks read the vault registry ~/.claude/vaults and the vault folders it lists, and write only inside the vault (_meta/index.json), a small per-session state file in the system temporary directory ($TMPDIR), and the plugin's own data directory; the init, seed, sync, tidy, trace and release procedures run the vault CLI through the terminal and commit to the vault's own git repository, and vault init appends the vault to ~/.claude/vaults and, only when you pass its opt-in flags, copies a vault launcher to ~/.local/bin (--install-cli), installs report-only post-merge and pre-push hooks in your code checkouts (--git-hook), or adds a post-commit hook that pushes every vault commit to the remote you name (--backup-remote); vault export write
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 lightkeeper
Said no? Turn it on later:
hermes plugins enable lightkeeper
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 S3-compatible backup and restore operations for Hermes
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install s3
Said no? Turn it on later:
hermes plugins enable s3
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).
Preview, publish, and manage websites hosted on Showly from Hermes Agent, through the hosted Showly MCP server (Streamable HTTP). Portable Agent Plugins v1 package (mcp.json + skills/).
Good to know before installing
Tool calls go to Showly's hosted MCP server at mcp.showly.ai. The bundled showly-hosting skill makes Showly the default for any deploy, preview or share request that names no provider, and new links default to public (anyone with the link). The showly-reports skill runs a bundled local Node.js renderer (vendored answer-me-with-html 0.4.14, upstream self-update check removed) through the shell and uploads the report directory to Showly when sharing is requested. Sign-in is out of band: `npx @showly/mcp-server login --to hermes` (unpinned) writes MCP_SHOWLY_API_KEY into ~/.hermes/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 showly
Said no? Turn it on later:
hermes plugins enable showly
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).
Official Hermes Agent integration for Frihet — discoverability, setup, diagnostics, and operating guidance for the Frihet MCP. Zero duplicated tools; the canonical Frihet MCP surface stays in @frihet/mcp-server and the remote endpoint https://mcp.frihet.io/mcp.
Good to know before installing
/frihet setup and /frihet doctor send an unauthenticated MCP initialize to mcp.frihet.io (or FRIHET_MCP_URL); /frihet doctor also makes one read-only list_invoices call through Hermes's MCP client once you grant plugins.entries.frihet.mcp_allowlist; reads ~/.hermes/config.yaml only to check that an mcp_servers.frihet block exists; the pre_tool_call hook prompts before Frihet MCP calls it classifies as irreversible or unknown, but reads and draft-first creates pass without a prompt unless the server sets 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 frihet
Said no? Turn it on later:
hermes plugins enable frihet
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).
GitHub PRs and issues as a right workspace pane in Hermes Desktop, via the connected gh CLI.
Good to know before installing
Shells out to your logged-in gh CLI; the in-pane update pill only runs `hermes plugins update githermes` (no fetch-and-replace) and currently compares against upstream main rather than the pinned sha.
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 githermes
Said no? Turn it on later:
hermes plugins enable githermes
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.
Tracekit
Community
Tracekit's signer decides every tool call before it runs (deny blocks, ask waits for a person) and signs each call and its outcome into a run you can verify offline.
Tracekit's signer decides every tool call before it runs (deny blocks, ask waits for a person) and signs each call and its outcome into a run you can verify offline.
Good to know before installing
Every tool call's name and arguments, plus its result (in full for tools Tracekit's policy marks untrusted, otherwise a SHA-256 digest), go to a Tracekit signer: by default a same-user dev signer (`python -m tracekit signer serve --dev`) that the plugin starts on first use as a detached background process, which keeps running after Hermes exits until 15 minutes idle and stores signed runs plus each call's redacted arguments and results (encrypted, kept 7 days) under your user's data directory (~/.local/share/tracekit on Linux), or a remote signer when you set TRACEKIT_SIGNER to an https:// URL. The plugin fails closed: when tracekit-ai can't load or the signer refuses or can't be reached, tool calls are blocked (including every tool an execute_code script calls), and an `ask` decision holds the call until someone approves it in Tracekit or Hermes's hook timeout runs out.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs (this one needs TRACEKIT_SIGNER_TOKEN), then asks whether to turn it on.
hermes plugins install tracekit
Said no? Turn it on later:
hermes plugins enable tracekit
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.
Kanban Gantt
Community
One board, three views, and each answers a different question: List says what the work is, Timeline says when it happened, Graph says what hangs off what.
One board, three views, and each answers a different question: List says what the work is, Timeline says when it happened, Graph says what hangs off what. The Graph view draws a card per task and an arrow per dependency, laid out top-down by longest path (or left to right, by status, or as a compact grid), with a transitive reduction that counts the arrows it implies instead of hiding them, links you draw, move and delete where they are drawn, and a click that lights the whole upstream and downstream lineage of a task. The Timeline renders any board as real work timestamps (created/started/review/done), with a collapsible tree column (boxed fold controls, tree connectors, per-board fold state, multi-parent links kept visible), live-activity arcs, filters, bulk actions, and a period selector from 6 hours to 7 days. The List view trades the bars for a dense table — task, assignee, status, last activity. Quick scopes (Running, Blocked, Open, All) narrow the board in all three views alike, and a lineage switch widens any of them to what leads to the selection and what follows it. Tasks are created from the page (project, workspace, skills, model override, goal mode), re-parented by drag & drop or from the drawer, and moved between boards with their children; any task with files carries a paperclip, its files listed on demand — the board payload carries only a count — and previewed by type (markdown, plain text, image, PDF, HTML in a sandboxed frame) or downloaded. Zero API keys, zero model tokens; live updates follow the plugin's /events websocket, which a gateway turns off with KANBAN_GANTT_WS=0 (the page then polls every 60 s and shows a Refresh button).
Good to know before installing
Desktop page plus dashboard routes that read and write the local kanban boards (create tasks with workspace path, skills, model override and goal mode; re-parent; move a task, with its children, to another board — the copy is verified there before the source rows and their links are deleted; edit status, assignee and description; bulk archive and delete; read task attachment files from the board's own attachments folder for preview and download); live updates over a dashboard-authenticated /events websocket, off with KANBAN_GANTT_WS=0; no third-party network.
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 kanban-gantt
Said no? Turn it on later:
hermes plugins enable kanban-gantt
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.
Agent Mixer
Community
A mixing desk for how much a coding agent may change: four enforced dials (edits, write-scope globs, per-turn line budget, new-dependency policy) with /mixer, live presets, a Desktop dial panel in the composer, a…
A mixing desk for how much a coding agent may change: four enforced dials (edits, write-scope globs, per-turn line budget, new-dependency policy) with /mixer, live presets, a Desktop dial panel in the composer, a web-dashboard saved-defaults bar and a steering-console page. Also drives Claude Code, Codex and Gemini CLI through the agent-mixer hook command.
Good to know before installing
Enforces edit/scope/line-budget/new-dependency dials in pre_tool_call (blocks, or escalates "ask" to Hermes approval, which /yolo skips) and adds a one-line rules context via pre_llm_call when a dial is off its default; per-session dial state is stored under HERMES_HOME/plugin-data/hermes-agent-mixer/sessions/, and the Desktop panel and dashboard bar write the plugin's own saved defaults into config.yaml (plugins.entries.hermes-agent-mixer.settings); the separately installed agent-mixer CLI keeps its own state in ~/.config/agent-mixer and ~/.local/state/agent-mixer, writes hook entries into ~/.claude/settings.json, ~/.claude/commands/mixer.md, ~/.codex/hooks.json or ~/.gemini/settings.json only when run as `agent-mixer install <harness> --yes`, and `agent-mixer ui` serves a token-guarded console on 127.0.0.1:8377.
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-agent-mixer
Said no? Turn it on later:
hermes plugins enable hermes-agent-mixer
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).
Field notes for a Hermes install: search known pitfalls before debugging and record new ones with root cause; register every local core edit as a patch note with a declarative checks file, then verify read-only after each hermes update (OK / MISSING / UPSTREAMED / N/A / UNKNOWN with reason) and list core files changed outside the registry. Ships 42 pitfalls and 10 patch recipes checked against 0.22.0. Optional pinned Telegram dashboard edited in place by an hourly no_agent cron job. Portable Agent Plugins v1 package (plugin.json + skills/); no tools, hooks, middleware or env vars; never edits the core or runs patch scripts.
Good to know before installing
Its scripts only read the Hermes core (pyproject.toml, read-only git status/diff) and never edit it or run patch scripts; the skill does walk the agent through hand-editing core files from its patch recipes and re-running the owner's re-apply scripts, each only after the user says yes. Network: Telegram Bot API only when the dashboard is enabled (reuses the agent's TELEGRAM_BOT_TOKEN from env or ~/.hermes/.env), and the public GitHub API without credentials (latest release plus compare against the installed commit SHA, cached 12 h; upstream.check=false disables). Writes: the notes store (~/.hermes/field-notes), ~/.hermes/field-notes.json, state and lock files in ~/.hermes/cache, a copy of the cron script in ~/.hermes/scripts when install.py runs (the skill runs it on first use), and one hourly no_agent cron job only when the user runs install_cron.py (it reads ~/.hermes/installs/*/facts.json to find Hermes' own Python and re-runs itself with it). Without a dashboard chat, install_cron.py sends the job's alerts to the Telegram home channel when TELEGRAM_HOME_CHANNEL is set (read from env or ~/.hermes/.env), else local.
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-field-notes
Said no? Turn it on later:
hermes plugins enable hermes-field-notes
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 boards for UI mockups. The agent builds a board from a decisions.json file and your prototype pages; you open it on desktop or phone, mark each decision approve, change or skip, pin comments on the mocks and press Send, and your agent wakes up with the feedback as a markdown digest. Four mock_review tools, the `hermes mockreview` CLI and slash command, and a bundled usage skill. Boards are served from your computer (loopback by default, or your tailnet) or published on here.now.
Good to know before installing
When autostart is on (off until `hermes mockreview setup`), a board web server (127.0.0.1:8790 by default) and one worker thread run inside the gateway process; setting host to 0.0.0.0 exposes every review to your network with no login. Outbound calls: a signed POST to the local Hermes webhook (localhost:8644) on Send, and, only for reviews hosted on here.now, https://here.now/api/v1 with your HERENOW_API_KEY. Secrets are read through Hermes's secret scope only. Runs `bash publish.sh` from the official here-now skill (key in the child environment), `tailscale status --json` and `tailscale serve status --json` (read only; on macOS the Tailscale app's own binary if the symlink crashes) for tailnet hosting, and from `setup` the hermes subcommands `--version`, `gateway status`, `config get/set/unset`, `webhook subscribe`, `webhook remove` (only for a stale self-delivering route, after you confirm) and `plugins validate`. Reviews are stored under $HERMES_HOME/workspace/mock-reviews. No telemetry, no update check, no Python dependencies.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs (this one needs HERENOW_API_KEY, MOCKREVIEW_WEBHOOK_SECRET), then asks whether to turn it on.
hermes plugins install mock-review
Said no? Turn it on later:
hermes plugins enable mock-review
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).
MIT-licensed product judgment across User, Evidence and Business flows, with one portable Agent Plugins v1 skill and all 91 original skill files.
Good to know before installing
Skill instructions run bundled local Python/jsonschema validators and read/write user-controlled local briefs, context packs and review outputs, which can contain personal data. Optional evidence gathering uses explicitly user/host-authorized tools, including external search and connectors; selected claims, queries or evidence go to the chosen services under their policies. There is no publisher-operated receiving endpoint, retention service, telemetry or background process. The bundled jev_gate.py runs offline; connecting a remote decision service needs separate configuration and consent. No MCP server, native plugin entrypoint, registered tools, hooks or middleware.
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 ux3-product-design-harness
Said no? Turn it on later:
hermes plugins enable ux3-product-design-harness
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).
Shodan tools with complete capability declarations.
How to install
Install it. Hermes asks for any API keys it needs (this one needs SHODAN_API_KEY), then asks whether to turn it on.
hermes plugins install shodan
Said no? Turn it on later:
hermes plugins enable shodan
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.
Redline
Community
Guided Markdown and source-code review in Hermes Desktop — annotate passages or line-numbered code in a docked pane and send comment/question/change-request batches back to the agent.
Guided Markdown and source-code review in Hermes Desktop — annotate passages or line-numbered code in a docked pane and send comment/question/change-request batches back to the agent; cards are offered automatically on delivered Markdown.
Good to know before installing
Reads any user-selected local .md/.markdown or source-code path (.py .sh .js .ts .tsx .c .h .cpp .go .rs .rb .yaml .yml .json .toml .sql, so config files such as config.yaml or auth.json are shown if you open them) up to 1 MB (symlink-resolved), writes only a <file>.redline.json sidecar beside it, and in Hermes Desktop chats the transform_llm_output hook appends a ::redline directive line to qualifying final agent responses; syntax highlighting uses Pygments in the local gateway host with plain-text fallback; no network beyond the local gateway, no credentials, no telemetry, no subprocesses; the pane polls the local backend every 3 s while open.
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 redline
Said no? Turn it on later:
hermes plugins enable redline
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.
Ample
Community
Ample (ample.computer) as native tools: deploy the websites and web apps Hermes builds to public HTTPS URLs, release changes to the same URLs, read build and runtime logs, list and delete apps.
Ample (ample.computer) as native tools: deploy the websites and web apps Hermes builds to public HTTPS URLs, release changes to the same URLs, read build and runtime logs, list and delete apps; bundled skill ample:web-apps.
Good to know before installing
Maintained by Ample, the hosting service it deploys to; talks to Ample's API (https://api.ample.computer); on first use downloads the pinned Ample CLI 0.1.36 from Ample's release server (https://get.ample.computer/releases/0.1.36/, SHA-256 checked, never self-updated; cli_path uses your own binary instead) and runs it as a subprocess with an allowlisted environment and its crash reports off; ample_deploy uploads the folder it is given (home, Hermes home and credential directories refused) and writes ample.toml and .ample/ into it; with no AMPLE_CLIENT_ID/AMPLE_CLIENT_SECRET or AMPLE_TOKEN set, the first deploy signs up for a free Ample account (deleted after two days unless claimed) and saves its credential in Hermes's .env. A pre_tool_call hook sends every delete (scoped to that app) and the first publish of each folder to Hermes's approval gate; only folders the user has published, recorded in plugin state, skip it. No telemetry, no 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 ample
Said no? Turn it on later:
hermes plugins enable ample
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.
Session Flow
Community
Session pane for Hermes Desktop: Firefox-style session groups with project grouping, list/grid views, Ctrl+Scroll session cycling with a HUD, session-row styling and a built-in settings page.
Session pane for Hermes Desktop: Firefox-style session groups with project grouping, list/grid views, Ctrl+Scroll session cycling with a HUD, session-row styling and a built-in settings page. All timers are ctx-owned intervals disposed on unload (live-session poll ~30 s, session refresh ~45 s, project cache 60 s, activity expiry 30 s). The standalone full build in this repo (full/plugin.js, install.sh) is not part of this catalog entry. Contact: info@agantila.com.
Good to know before installing
SDK-only Desktop plugin: reads the local Hermes gateway through public RPCs (projects.tree, projects.list, session.list, session.active_list, session.context_breakdown) and, on your actions, writes through public RPCs (projects.create/update/add_folder/remove_folder/set_primary/set_active/delete, session.create, session.cwd.set, session.title, session.branch_stored, session.archive, session.close, session.delete); captures Ctrl+mouse-wheel across the app (browser zoom suppressed except over canvases, editors and zoomable images) to cycle sessions; settings and groups stay in the app's local plugin storage; no outbound network, no shell commands, no credentials, no self-update.
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 session-flow
Said no? Turn it on later:
hermes plugins enable session-flow
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.
Gh Ref
Community
@gh:owner/repo#123 context references: attach a public GitHub issue or pull request to a message (title, state, labels, description, the latest comments; for pull requests also the branches, merge state and change size).
@gh:owner/repo#123 context references: attach a public GitHub issue or pull request to a message (title, state, labels, description, the latest comments; for pull requests also the branches, merge state and change size).
Good to know before installing
Sends anonymous HTTPS GET requests to api.github.com only (the referenced owner/repo and number leave the machine; no token, public repositories only, 60 requests an hour per IP address); no files written, no subprocess; register() only adds the @gh: reference provider.
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 gh-ref
Said no? Turn it on later:
hermes plugins enable gh-ref
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).
Per-provider quota and balance readout in the Hermes desktop status bar. Per-provider chips with eye toggles, a popover with every quota window, reset times, balance and key-cap amounts, near-limit amber tint, a desktop notification per threshold crossing, and burn-rate projections (ETA to reset or depletion). Providers: OpenCode Go, OpenAI Codex, OpenRouter, plus Anthropic, GitHub Copilot, Nous, Z.AI, Kimi, MiniMax and DeepSeek when configured.
Good to know before installing
Resolves provider credentials read-only (Hermes account-usage helpers and profile secrets, else the profile's .env and the vendor CLI login files ~/.codex/auth.json, ~/.claude/.credentials.json and ~/.local/share/opencode/auth.json) and sends each only to that provider's usage endpoint; the GitHub Copilot check presents the VS Code / Copilot client identity. No subprocesses, 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 provider-usage
Said no? Turn it on later:
hermes plugins enable provider-usage
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.
B64 Ref
Community
@b64: context references that decode encoded text for the model: @b64:<value> takes base64 (standard or URL-safe, padded or not), hex (0x48656c6c6f, 48:65:6c, \x48\x65), percent-encoding or a data: URI, told apart by…
@b64: context references that decode encoded text for the model: @b64:<value> takes base64 (standard or URL-safe, padded or not), hex (0x48656c6c6f, 48:65:6c, \x48\x65), percent-encoding or a data: URI, told apart by the value's shape or picked with base64:, base64url:, hex: or url:, and attaches what it holds: text as text with control and invisible characters escaped, JSON pretty-printed, gzip and zlib inflated in memory (at most 1 MiB, one layer), and binary named by its format (PNG and GIF with their size, JPEG, PDF, ZIP, ELF, SQLite and more) with a hex dump of its first 128 bytes; notes point out a value that was cut short, padding that is off, a second layer of encoding and a JWT inside; the TUI and Desktop composer preview a value as you type it.
Good to know before installing
Reads only the text of the reference itself, using Python's base64, json, urllib.parse and zlib; nothing decoded is run, opened or saved; no network, no files read or written, no subprocess, no credentials read; the plugin keeps no copy of what it decodes, but the @b64: reference and the decoded contents become part of the message sent to your model provider and saved in the session history, so decode only values you are willing to share with it; register() only adds the @b64: reference provider.
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 b64-ref
Said no? Turn it on later:
hermes plugins enable b64-ref
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.
BackupDrill
Community
Back up a Supabase project to your own S3, R2 or B2 bucket with the open-source backupdrill CLI (the Postgres database, plus Storage files once their S3 keys are set), then restore the snapshot into a throwaway…
Back up a Supabase project to your own S3, R2 or B2 bucket with the open-source backupdrill CLI (the Postgres database, plus Storage files once their S3 keys are set), then restore the snapshot into a throwaway Postgres in Docker to check it comes back.
Good to know before installing
A skill only, with no Python, hooks or server: it has the agent run npx -y backupdrill@2.7.1 in shell commands, which npx downloads from npm on first run. The CLI connects to the user's Supabase project and their own bucket, and a drill pulls a postgres image from Docker Hub. The connection string and bucket keys stay in a .env.backupdrill file the user fills in; each command loads it in a subshell and never prints it. Nothing is sent to BackupDrill.
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 backupdrill
Said no? Turn it on later:
hermes plugins enable backupdrill
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.
Progress Squash
Community
Collapses carriage-return progress bars (tqdm, git clone --progress and similar) in terminal output to the frame a terminal would show, so the model reads one line instead of every redraw: a 300-step tqdm loop goes…
Collapses carriage-return progress bars (tqdm, git clone --progress and similar) in terminal output to the frame a terminal would show, so the model reads one line instead of every redraw: a 300-step tqdm loop goes from 17,396 characters to about 120. Other text written between redraws, CRLF line endings and output without a carriage return are left as they are.
Good to know before installing
Register() adds one transform_terminal_output hook that rewrites the output string in memory; no tools, network access, subprocesses, files, settings or environment variables.
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 progress-squash
Said no? Turn it on later:
hermes plugins enable progress-squash
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.
Stream Speed
Community
How fast each model starts answering and how fast it writes: /speed shows the median and p90 time to first text and the median characters per second per model, provider or day, or lists the newest streams (ten, or up…
How fast each model starts answering and how fast it writes: /speed shows the median and p90 time to first text and the median characters per second per model, provider or day, or lists the newest streams (ten, or up to 100 with /speed recent 50), for the last 24 hours, any number of hours or days, the latest session or everything recorded, from the streaming hooks.
Good to know before installing
Writes per-response stream timings (ids, model, timings, character count, no text) to plugin-data/stream-speed/streams.jsonl, trimmed to 5,000 rows past 2 MB; atexit waits up to 1 s; no network
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 stream-speed
Said no? Turn it on later:
hermes plugins enable stream-speed
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).
Reduce Hermes tool-schema overhead and show selector savings in the dashboard.
Good to know before installing
Optional mode jev (off by default; needs TYPESAFE_API_KEY) sends each turn's request text and up to 32 candidate tool names and descriptions to TypeSafe (api.typesafe.ai); all other modes run locally. Optional user-run scripts can patch Hermes core on older Hermes and install a systemd user self-heal unit; the dashboard advisor edits the tool_slimmer section of config.yaml with a 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 tool-slimmer
Said no? Turn it on later:
hermes plugins enable tool-slimmer
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).
Durable project ownership and bounded initiative management for Hermes fleets.
Good to know before installing
A `pre_llm_call` hook adds a project's saved rules, decisions and facts to turns of bots assigned to that project; objective checks run allowlisted local programs only at autonomy level 3+ and GitHub objectives shell out to your own `gh api`; it creates and edits Hermes Kanban tasks through Hermes' internal kanban modules (title/body/tenant written directly to kanban.db), keeps its own SQLite DB under `<Hermes home>/plugin-data/hermes-dockyard/`, and its optional HTTP API (`/stewardship/v1`, 127.0.0.1) stays closed until you set a bearer token.
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-project-stewardship
Said no? Turn it on later:
hermes plugins enable hermes-project-stewardship
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).
BM25-based skill retrieval for large skill libraries. An llm_request middleware rewrites the outgoing request's <available_skills> block to names only (the same bytes every turn, so prompt caching is unaffected), and a pre_llm_call hook appends the top-K matching skill descriptions to the current user message. Uses public plugin surfaces only (register_middleware, register_hook); reads SKILL.md frontmatter from the profile's skill roots and imports a few agent.prompt_builder / agent.skill_utils helpers read-only. pyyaml is the only dependency.
Good to know before installing
Optional Jev rerank, off by default: with SKILL_RETRIEVAL_RERANK=jev and TYPESAFE_API_KEY set (read from the environment only), each turn POSTs the user message (injection patterns stripped, capped at 2000 chars) and up to 12 shortlisted skill descriptions (400 chars each) to api.typesafe.ai (3 s timeout, no retries, BM25 order kept on error), and appends a local JSONL log (timestamp, session id, 200-char query excerpt, both orders) under $HERMES_HOME/data/jev-trial/, rotated past 5 MB. With the rerank off it makes no network calls and writes no files.
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 skill-retrieval
Said no? Turn it on later:
hermes plugins enable skill-retrieval
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).
Parallel AI task orchestrator. Paste a task list into the desktop panel; Hermes phases it into dependency-ordered groups through the agent_batch_plan / agent_batch_phase tools (or you group it yourself in one click), then each phase dispatches one independent coding agent per task to a GitHub Actions workflow in a repo you name — own branch, own PR — and the panel tracks the runs and the PRs. GITHUB_TOKEN is the only required credential; AGENT_BATCH_REPO, AGENT_BATCH_WORKFLOW, AGENT_BATCH_MODEL and AGENT_BATCH_STATE are optional (the last is a directory override for the plan file, not a secret) and each can be set per plan in the panel instead.
Good to know before installing
Hermes can read and rewrite the stored phase plan through the agent_batch_plan / agent_batch_phase tools, but dispatch is always a confirmed click in the panel. Each dispatch fires a GitHub workflow_dispatch with your GITHUB_TOKEN (Actions + pull-requests write) on the repo you name. The workflow you copy into that repo installs third-party coding-agent CLIs (cline by default; the others are unpinned npm packages) and runs them unattended, with auto-approve / skip-permissions, repo write access and your provider secrets. They push real branches and open real PRs, and your task text and repository code go to that runner's model vendor. LLM cost belongs to the runner: cline spends Cline Credits or a BYOK key, opencode free models need its zen relay, kilo needs an active Kilo Go subscription (keyless CI runs fail), and claude-code / codex / dsh need their own keys.
From the plugin's catalog listing.
How to install
Install it. Hermes asks for any API keys it needs (this one needs GITHUB_TOKEN), then asks whether to turn it on.
hermes plugins install agent-batch
Said no? Turn it on later:
hermes plugins enable agent-batch
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).
Today's git digest in the current repo, plus the last failed terminal command the Hermes agent ran. Desktop pane, status chip, optional local LLM summary. Does not watch OS or editor terminals.
How to install
Install it. Hermes asks for any API keys it needs, then asks whether to turn it on.
hermes plugins install debug-desk
Said no? Turn it on later:
hermes plugins enable debug-desk
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.
Game Editor
Community
In the Godot 4 project open in your editor, it places assets into the scene, plays the game, and the next plan fixes each error from its file and line.
In the Godot 4 project open in your editor, it places assets into the scene, plays the game, and the next plan fixes each error from its file and line.
Good to know before installing
Needs Godot 4.7 or later, the Godot AI 4.x add-on, and uv. The first start downloads godot-ai==4.3.0 from pypi.org and files.pythonhosted.org. There is no API key. The editor is reached only on 127.0.0.1. A server this plugin starts sends no telemetry. The add-on follows godot_ai/telemetry_enabled (on by default) and can send usage to godot-ai-telemetry-v2-pudmurzsnq-uw.a.run.app. plan_change sends only the session list. apply_plan asks once. A plan that deletes, overwrites, drops unsaved work, writes outside the project or runs code it wrote is not made when what is lost exceeds 1,700 characters, and otherwise runs only after confirm_risky_plan. Play and tests ask once and run the project's code. Tools are on by default on every surface, including the gateway, so anyone who can message Hermes there can ask and approve. This plugin keeps no user list of its own. Cron, hermes chat -q and unattended runs decline. Data stays in <HERMES_HOME>/plugin-data/agent-plugin-game-editor-<id> after removal. Stop with hermes plugins disable game-editor, or hermes tools for one surface. macOS only.
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 game-editor
Said no? Turn it on later:
hermes plugins enable game-editor
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).
Move data across cloud storage and exchange end-to-end encrypted messages, files and media between agents with Beam Network (Bittensor SN105).
How to install
Install it. Hermes asks for any API keys it needs (this one needs BEAM_API_KEY), then asks whether to turn it on.
hermes plugins install beam
Said no? Turn it on later:
hermes plugins enable beam
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).
Hermes Desktop session analytics and provider-reported usage: annual heatmaps, token mix, model breakdowns and source-period summaries.
Good to know before installing
Reads the selected profile's state.db read-only and, for the Providers tab, an Activity cache under plugin-data/hermes-activity plus the optional Quota quota_cache.json. Fetch provider data runs a local Node.js (22.18+) collector and, only on that explicit click, calls chatgpt.com through Hermes's signed-in Codex request route (requires_auth: openai-codex) and api.commandcode.ai, opencode.ai and ollama.com with the optional API keys you configure as plugin secrets; it stores usage aggregates and hashed key identifiers locally. No other credential reads, telemetry or sign-in flow; remote profiles go through the configured gateway.
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-activity
Said no? Turn it on later:
hermes plugins enable hermes-activity
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).
Read-only project inspection and asset audits with playable-slice, asset-production, and verification skills.
Good to know before installing
Lists file names and sizes under any absolute project path the agent supplies (skipping hidden, symlinked and credential-named entries), reads that project's project.godot/export_presets.cfg and, on request, hashes asset files; no writes, no shell commands, no network calls.
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 game-dev
Said no? Turn it on later:
hermes plugins enable game-dev
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.
Project Manager
Community
Git workbench for the Hermes Dashboard and Desktop: discovers the git repositories under the folders you watch, and inspects or changes them per file — stage, unstage, diff and discard — plus commit, push, pull…
Git workbench for the Hermes Dashboard and Desktop: discovers the git repositories under the folders you watch, and inspects or changes them per file — stage, unstage, diff and discard — plus commit, push, pull, branches, commits and worktrees.
Good to know before installing
Reads the watch paths from its own SQLite table (under HERMES_HOME/plugin-data/project-manager) and git metadata (branch, remote, ahead/behind, commit subjects, status and diff text) from the repositories under them, and lists folder names under any path you type when adding a watch folder; file contents are read only as diff text returned to the requesting UI; the five agent tools are read-only. Writes on request only from the Dashboard/Desktop UI: git add, commit, push, stash (including stash -u before a branch switch), fetch, merge --ff-only (also as "pull all" across every watched repository), restore, clean -f, git switch, and git worktree remove (including --force and remove-all). Discard on a tracked file restores it from HEAD and drops the uncommitted edits; on a file in no commit (untracked or staged-new) it leaves the bytes on disk but nothing here can restore a version of it, and both sit behind a confirmation naming which one you are doing. The Desktop Chat button starts a new Hermes session whose working directory is that project's folder, and the project name and branch chip open the repository's remote URL in your system browser. The optional standalone server (proje
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 project-manager
Said no? Turn it on later:
hermes plugins enable project-manager
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.
