ircv3-mcp
An IRCv3 MCP server that enables agents to act as a mini IRC client: read channels as transcripts, send messages, reply to threads, add reactions, fetch history, and manage channel membership via MCP tools.
README
ircv3-mcp
An IRCv3 MCP server: a mini IRC client that agents drive over the Model Context Protocol. The agent reads channels as rendered chat transcripts, sends messages, replies to threads, adds reactions, fetches history, and manages channel membership — all through a standard set of MCP tools backed by a full IRCv3 connection with SASL authentication.
Install
Requires Node >= 20. Nothing to install — it runs through npx and always pulls the latest
published version.
Setup is two steps, both in the Quickstart below: first configure an IRC account (host, nick, SASL), then register the server with your agent. The agent has nothing to talk to until at least one account is configured.
Quickstart
Configure an account with the interactive wizard — run through npx, nothing to install:
npx -y ircv3-mcp@latest configure
It prompts for host, nick, SASL mechanism (one of PLAIN, EXTERNAL, or SCRAM-SHA-256), and a hidden password, and stores the password in your OS keychain.
Prefer a one-liner? add-account reads the password from stdin so it never hits shell history:
echo 'hunter2' | npx -y ircv3-mcp@latest add-account libera \
--host irc.libera.chat \
--nick mybot \
--sasl PLAIN \
--account mybot \
--channels '#test' \
--default \
--password-stdin
Verify connectivity:
npx -y ircv3-mcp@latest test libera
Then register the server with your agent and the tools listed below become available immediately:
claude mcp add ircv3-mcp -- npx -y ircv3-mcp@latest
Tools
Read-only
irc_list_networks— list all configured accounts and their connection statusirc_status— show connection status, nick, and active capabilities for an accountirc_read_history— fetch messages from a channel or DM as a rendered transcriptirc_list_conversations— list channels and DMs that had activity in a time windowirc_list_members— list current members of a channel with mode prefixesirc_whois— look up information about a nick
Writes
irc_send_message— send a message or multiline batch; supports threading viain_reply_toirc_react— add or remove an emoji reaction on a messageirc_join— join a channel, optionally with a keyirc_part— leave a channelirc_mark_read— advance the read marker for a conversation
Destructive
irc_redact— delete a message by msgid (cannot be undone)irc_send_raw— send a raw IRC protocol line (unrestricted)
Configuration
Accounts are stored in TOML at ~/.config/ircv3-mcp/config.toml. The XDG
XDG_CONFIG_HOME variable is respected; the path can also be overridden with
IRCV3_MCP_CONFIG_DIR.
Passwords are never stored in the config file, in tool output, or in logs. The server
stores them in the OS keychain (@napi-rs/keyring) and falls back to an AES-256-GCM
encrypted file (secrets.enc) when the keychain is unavailable. The encryption key comes from
a 0600 keyfile (secrets.key) or the IRCV3_MCP_SECRET_KEY environment variable.
See docs/config.example.toml for an annotated example.
Persistence
The server is stateless per session: it opens one IRC connection per configured account when
the agent session starts and uses draft/chathistory to catch up on messages since the last
read marker.
For gap-free 24/7 presence — keeping the connection alive between agent sessions and accumulating full history — point ircv3-mcp at a bouncer such as soju or ZNC, or an always-on server such as Ergo or ObbyIRCd.
Development
npm run ci # typecheck + lint + test + build (mirrors CI)
npm test # vitest unit tests only
npm run build # compile to dist/
Integration tests against a live Ergo instance are gated by IRC_IT=1 and are not yet wired
into the default test run. Set that variable to opt in when running locally against a
test server.
License
MIT
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.