mailmcp

Create draft

create_draft

Saves a draft in the account's Drafts folder. Nothing is sent; the owner reviews and sends it (or asks you to send_draft). Attachments: mailbox attachments, files from upload_attachment/request_upload, or inline content. The preferred way to prepare replies.

Input Schema

TableJSON Schema

Name	Required	Description
`cc`	No
`to`	Yes	Recipient addresses
`bcc`	No
`html`	No	Optional HTML (scripts stripped)
`text`	No	Plain-text body
`vars`	No	Values for its {{placeholders}}
`quote`	No	Quote the original (default true when replying)
`account`	Yes	list_accounts id
`subject`	No
`template`	No	list_templates name (paid)
`attachments`	No
`in_reply_to_uid`	No	uid being answered; prefer reply_draft / reply_send
`in_reply_to_folder`	No	Default: Gmail All Mail, else INBOX

Schema Changelog

Changes observed during successful MCP inspections.

Changed19 schema fields changedOct 5, 2026

changedInput schema / properties / account / description

Previous value: -"Account id from list_accounts"New value: +"list_accounts id"

removedInput schema / properties / attachments / description

Removed value: -"Files to attach: existing mailbox attachments, uploaded files (folder \"mailmcp-uploads\"), inline content, or local paths"

changedInput schema / properties / attachments / items / description

Previous value: -"Attach a mailbox attachment {uid, part[, folder, account]}, inline content {filename, content[, encoding]}, or a local file {path}"New value: +"{uid, part[, folder, account]} from a mailbox, {filename, content[, encoding]} or {path}"

removedInput schema / properties / attachments / items / properties / account / description
```
Removed value: -"Account holding the file (defaults to the sending account)"
```

removedInput schema / properties / attachments / items / properties / content / description

Removed value: -"File content, UTF-8 text or base64 (see encoding); for files the assistant writes itself"

removedInput schema / properties / attachments / items / properties / content_type / description
```
Removed value: -"MIME type of inline content, e.g. text/csv"
```
removedInput schema / properties / attachments / items / properties / encoding / description
```
Removed value: -"How `content` is encoded, default utf8"
```
removedInput schema / properties / attachments / items / properties / filename / description
```
Removed value: -"File name (required for inline content)"
```

removedInput schema / properties / attachments / items / properties / folder / description

Removed value: -"Folder of the source message (e.g. \"mailmcp-uploads\" for uploaded files)"

changedInput schema / properties / attachments / items / properties / part / description

Previous value: -"Attachment part id from get_message / list_uploads"New value: +"Part id from get_message / list_uploads"

changedInput schema / properties / attachments / items / properties / path / description

Previous value: -"Local file path (Claude Desktop / Claude Code only, within policy.attachment_dirs)"New value: +"Local path (stdio only, inside policy.attachment_dirs)"

removedInput schema / properties / attachments / items / properties / uid / description
```
Removed value: -"uid of the source message"
```

changedInput schema / properties / html / description

Previous value: -"Optional HTML body (scripts are stripped); when absent it is rendered from text"New value: +"Optional HTML (scripts stripped)"

changedInput schema / properties / in_reply_to_folder / description

Previous value: -"Folder/label path. Defaults to All Mail on Gmail, INBOX elsewhere."New value: +"Default: Gmail All Mail, else INBOX"

changedInput schema / properties / in_reply_to_uid / description

Previous value: -"uid of the message being answered: sets In-Reply-To/References, keeps the Re: subject and quotes the original. Prefer reply_draft / reply_send."New value: +"uid being answered; prefer reply_draft / reply_send"

changedInput schema / properties / quote / description

Previous value: -"Quote the original under the reply (default true when replying)"New value: +"Quote the original (default true when replying)"

addedInput schema / properties / template

Added value: +{
+  "description": "list_templates name (paid)",
+  "maxLength": 60,
+  "type": "string"
+}

addedInput schema / properties / vars

Added value: +{
+  "additionalProperties": {
+    "maxLength": 2000,
+    "type": "string"
+  },
+  "description": "Values for its {{placeholders}}",
+  "propertyNames": {
+    "pattern": "^[a-z_][a-z0-9_]{0,30}$",
+    "type": "string"
+  },
+  "type": "object"
+}

changedInput schema / required

Previous value: -[
-  "account",
-  "to",
-  "subject",
-  "text"
-]New value: +[
+  "account",
+  "to"
+]

Changed11 schema fields changedSep 22, 2026
- addedInput schema / properties / attachments / items / properties / uid / maxLength
```
Added value: +512
```
- removedInput schema / properties / attachments / items / properties / uid / maximum
```
Removed value: -9007199254740991
```
- addedInput schema / properties / attachments / items / properties / uid / minLength
```
Added value: +1
```
- removedInput schema / properties / attachments / items / properties / uid / minimum
```
Removed value: -1
```
- changedInput schema / properties / attachments / items / properties / uid / type
```
Previous value: -"integer"New value: +"string"
```
- addedInput schema / properties / attachments / maxItems
```
Added value: +20
```
- addedInput schema / properties / in_reply_to_uid / maxLength
```
Added value: +512
```
- removedInput schema / properties / in_reply_to_uid / maximum
```
Removed value: -9007199254740991
```
- addedInput schema / properties / in_reply_to_uid / minLength
```
Added value: +1
```
- removedInput schema / properties / in_reply_to_uid / minimum
```
Removed value: -1
```
- changedInput schema / properties / in_reply_to_uid / type
```
Previous value: -"integer"New value: +"string"
```
First observedSep 16, 2026

TDQS

A4.2/5.0

Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already declare a non-destructive write, but the description adds the crucial nuance that nothing is transmitted until the owner reviews and sends it, plus the permitted attachment sources. This is meaningful behavioral context beyond the annotations, though it omits any note on return values or permission/scope requirements.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Four short sentences, all front-loaded and each earning its place: purpose, send behavior, attachment sourcing, reply preference. No filler or redundancy.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 13-parameter write tool with no output schema, the description covers purpose, non-sending behavior, attachment sources, and reply intent adequately. It could say more about the account requirement and reply-context parameters, but nothing critical to a correct call is missing.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

With 69% schema coverage and nested attachment objects, the schema carries most of the burden. The description adds the tool names that feed attachments (upload_attachment/request_upload) but says nothing about the reply parameters (in_reply_to_uid, quote, vars) or how they interact, so it only marginally extends the schema.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource with location ("Saves a draft in the account's Drafts folder") and immediately distinguishes the outcome from its siblings ("Nothing is sent... or asks you to send_draft"). An agent can tell this apart from send_draft and reply_draft without opening a schema.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Gives clear context ("The preferred way to prepare replies") and names the alternative for sending (send_draft). It does not, however, address when to prefer reply_draft/reply_send for replies, so the routing is only partially complete.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

All mailmcp tools Inspect

Try in Browser

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Create draft

Input Schema

Schema Changelog

TDQS

Glama MCP Gateway

Resources