Sendmux Skills
Connect agent email workflows to Sendmux mailboxes, attachments, CLI, and MCP.
Sendmux’s complete canonical pack covers onboarding, sending, mailbox agents, attachments, management, MCP setup, and efficient tool use. Separate plugin bundles distribute the same Skills to supported hosts.
Use cases
- Set up agent inbox and sending workflows
- Manage messages and attachments
- Connect Sendmux CLI or MCP
How to use
Follow Sendmux account and tool setup before using a mailbox. Provide the intended email operation and choose CLI or MCP according to the upstream workflow.
Skills
agent-email-inbox
Use when an AI agent needs read-only triage of an existing Sendmux inbox or unsent reply text from its messages. Trigger for bounded inbox reviews, unread invoice or support searches, and draft-only replies, including agent inbox requests without the Sendmux name. Mailbox mutations and sending are outside this workflow.
sendmux/skillsemail-for-ai-agents
Use when an AI agent needs to send one outbound email through Sendmux with human approval, including a status notification or delegated agent send. Trigger when preparing the exact message for approval, executing an approved send, or reconciling an uncertain retry. Use only for a single-message workflow; bulk and batch sends are outside its scope.
sendmux/skillssendmux-attachments
Use Sendmux attachment workflows without wasting model context on base64. Use when uploading, downloading, reading, forwarding, or sending email attachments through Sendmux MCP, CLI, SDKs, or direct HTTP, especially when choosing file_path vs presigned upload URL vs inline base64, reading inbound attachments with mailbox_read_attachment, fetching short-lived download_url links, or attaching local files to outbound mail.
sendmux/skillssendmux-cli
Use when a user wants to run the Sendmux **CLI** (the `sendmux` terminal binary) — installing it, OAuth login/logout, connection checks, agent inbox registration and owner invites, managing API-key or OAuth profiles, key-scope preflight errors (e.g. root vs mailbox vs agent key), `--json` output, or invoking Management, Mailbox, and Sending operation commands (domains, mailboxes, webhooks, messages, search, sending email, attachments, delivery logs) via `sendmux COMMAND --flag` syntax. Trigger for "what sendmux command", "how do I run/install sendmux", CLI flag or error questions, and profile/credential setup in the terminal. Do NOT trigger for using Sendmux through MCP tools, the REST API directly from code, or general email-sending strategy questions unrelated to the CLI.
sendmux/skillssendmux-email-for-agents
Use when a user wants to design or plan how an AI agent gets and uses email — giving an agent its own inbox, deciding between owner-managed vs self-registered provisioning, building a receive/triage/reply loop, sending notifications with human approval, or choosing which credential/surface to use — especially when they're unsure how to architect this, even if Sendmux isn't named. Do NOT use for a user who already knows their setup and just wants a specific low-level command, code snippet, or API payload (e.g., "write the JSON for an attachment", "install the CLI", "create a mailbox key from the terminal") — route those directly to CLI/API syntax instead of this architecture/routing skill.
sendmux/skillssendmux-getting-started
Use when a user needs Sendmux setup, OAuth connection checks, agent inbox registration, owner linking, credential validation, surface selection, or a first harmless call through MCP, CLI, or an SDK.
sendmux/skillssendmux-mailbox-agent
Operate inside an already-existing Sendmux mailbox on behalf of an AI agent: read, search, count, sync, triage, file, delete, thread, save or edit a reply draft, or send a message using an API key, scoped agent token, or authorised OAuth profile. Trigger when the user wants an agent to inspect an inbox, find or mark messages, resume a prior sync state, prepare a human-editable draft, or send from that mailbox's identity. Do NOT use this for creating, provisioning, or administering mailboxes, domains, or API keys themselves — that setup and admin work belongs to sendmux-management, not this skill.
sendmux/skillssendmux-management
Use this skill for Sendmux account/team administration with a root API key or Management-scoped OAuth connection. This is the go-to skill for domain lifecycle work — create a sending domain, fetch its DNS/zone-file records, verify DNS/MX — even if asked as a single quick step. Also use it for provisioning, suspending, or deleting mailboxes and mailbox API keys; configuring sending accounts/providers; creating, testing, or rotating webhooks; checking billing, spend, or balance; and pulling delivery, incoming, or metrics logs for reporting or admin purposes. This is account/team-level administration, not inbox usage: do not trigger for reading, searching, triaging, syncing, or replying to messages inside a specific mailbox (use sendmux-mailbox-agent), one-off transactional sending (sendmux-send-email), or initial setup/login (sendmux-getting-started).
sendmux/skillssendmux-mcp-setup
Set up, configure, and troubleshoot Sendmux Model Context Protocol (MCP) servers for AI agent clients. Use for installing sendmux-mcp, wiring the hosted OAuth endpoint, running local stdio/HTTP servers, setting mailbox/management/sending key scopes, adding bearer headers, and writing MCP server config (JSON/TOML) for Claude Code, Cursor, Codex, VS Code/Copilot, Copilot CLI, Gemini CLI, Cline, or Windsurf/Cascade. Also use for diagnosing MCP OAuth/DCR problems: dynamic client registration missing a `resource` parameter, `invalid_target` errors, or mismatches between the client's authorisation request, the granted scope, and the token audience/resource binding for `https://mcp.sendmux.ai/mcp`. Do not use this skill just because a client is already sending, searching, or reading email through Sendmux MCP tools — that is ordinary tool use, not server setup, connection, or auth configuration.
sendmux/skillssendmux-send-email
Use when a user needs advice or implementation help for sending one or many emails through Sendmux: choose the HTTP Sending API versus SMTP, design single or batch sends, attach files, apply idempotency, or inspect send results through MCP, CLI, SDK, or direct HTTP. Not for initial credential, API-key, OAuth, or profile setup—use sendmux-getting-started; not for reading, searching, syncing, or triaging an inbox—use sendmux-mailbox-agent.
sendmux/skillssendmux-token-efficient-usage
Use when the question is about *how to call* Sendmux cheaply, not what to send: choosing batch vs. looped sends, mailbox search/count/batch-read vs. full fetches, sync deltas, cursor pagination, ETags/conditional requests, idempotency keys, or picking between MCP, the sendmux CLI, an SDK, or direct HTTP. Trigger for reducing token/round-trip/API-call cost, avoiding broad mailbox or log fetches, or optimising a polling/sync loop. This is a routing/cost-decision skill, never a content-authoring one. If the request asks to write, draft, compose, or format the actual JSON body, subject, text, or attachment bytes of an email — even a single email, even one with an attachment, even inside a larger batch — do NOT use this skill; route that to sendmux-send-email or sendmux-attachments instead. Ask yourself: is the user asking "what's the cheapest way to call Sendmux," or "what should this email/attachment contain"? Only the former belongs here.
sendmux/skills