Skip to main content
Glama

create_draft

Create draft emails in Gmail by specifying recipients, subject, and body content, or use raw RFC 2822 format for complete message control.

Instructions

Create a draft email in Gmail. Note the mechanics of the raw parameter.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
rawNoThe entire email message in base64url encoded RFC 2822 format, ignores params.to, cc, bcc, subject, body, includeBodyHtml if provided
threadIdNoThe thread ID to associate this draft with
toNoList of recipient email addresses
ccNoList of CC recipient email addresses
bccNoList of BCC recipient email addresses
subjectNoThe subject of the email
bodyNoThe body of the email
includeBodyHtmlNoWhether to include the parsed HTML in the return for each body, excluded by default because they can be excessively large

Implementation Reference

  • The core handler function for the 'create_draft' tool. It handles OAuth setup via handleTool, constructs raw email if needed using constructRawMessage, calls Gmail API to create draft, processes the response payload, and formats it for MCP.
    async (params) => {
      return handleTool(config, async (gmail: gmail_v1.Gmail) => {
        let raw = params.raw
        if (!raw) raw = await constructRawMessage(gmail, params)
    
        const draftCreateParams: DraftCreateParams = { userId: 'me', requestBody: { message: { raw } } }
        if (params.threadId && draftCreateParams.requestBody?.message) {
          draftCreateParams.requestBody.message.threadId = params.threadId
        }
    
        const { data } = await gmail.users.drafts.create(draftCreateParams)
    
        if (data.message?.payload) {
          data.message.payload = processMessagePart(
            data.message.payload,
            params.includeBodyHtml
          )
        }
    
        return formatResponse(data)
      })
    }
  • Zod schema defining the input parameters for the create_draft tool, including optional raw message, threadId, recipients, subject, body, and includeBodyHtml flag.
    {
      raw: z.string().optional().describe("The entire email message in base64url encoded RFC 2822 format, ignores params.to, cc, bcc, subject, body, includeBodyHtml if provided"),
      threadId: z.string().optional().describe("The thread ID to associate this draft with"),
      to: z.array(z.string()).optional().describe("List of recipient email addresses"),
      cc: z.array(z.string()).optional().describe("List of CC recipient email addresses"),
      bcc: z.array(z.string()).optional().describe("List of BCC recipient email addresses"),
      subject: z.string().optional().describe("The subject of the email"),
      body: z.string().optional().describe("The body of the email"),
      includeBodyHtml: z.boolean().optional().describe("Whether to include the parsed HTML in the return for each body, excluded by default because they can be excessively large")
    },
  • src/index.ts:229-263 (registration)
    MCP server.tool registration for 'create_draft', including name, description, input schema, and inline handler function.
    server.tool("create_draft",
      "Create a draft email in Gmail. Note the mechanics of the raw parameter.",
      {
        raw: z.string().optional().describe("The entire email message in base64url encoded RFC 2822 format, ignores params.to, cc, bcc, subject, body, includeBodyHtml if provided"),
        threadId: z.string().optional().describe("The thread ID to associate this draft with"),
        to: z.array(z.string()).optional().describe("List of recipient email addresses"),
        cc: z.array(z.string()).optional().describe("List of CC recipient email addresses"),
        bcc: z.array(z.string()).optional().describe("List of BCC recipient email addresses"),
        subject: z.string().optional().describe("The subject of the email"),
        body: z.string().optional().describe("The body of the email"),
        includeBodyHtml: z.boolean().optional().describe("Whether to include the parsed HTML in the return for each body, excluded by default because they can be excessively large")
      },
      async (params) => {
        return handleTool(config, async (gmail: gmail_v1.Gmail) => {
          let raw = params.raw
          if (!raw) raw = await constructRawMessage(gmail, params)
    
          const draftCreateParams: DraftCreateParams = { userId: 'me', requestBody: { message: { raw } } }
          if (params.threadId && draftCreateParams.requestBody?.message) {
            draftCreateParams.requestBody.message.threadId = params.threadId
          }
    
          const { data } = await gmail.users.drafts.create(draftCreateParams)
    
          if (data.message?.payload) {
            data.message.payload = processMessagePart(
              data.message.payload,
              params.includeBodyHtml
            )
          }
    
          return formatResponse(data)
        })
      }
    )
  • Helper function to construct the base64url-encoded raw RFC2822 email message from parameters, including thread headers and quoted content if replying to a thread.
    const constructRawMessage = async (gmail: gmail_v1.Gmail, params: NewMessage) => {
      let thread: Thread | null = null
      if (params.threadId) {
        const threadParams = { userId: 'me', id: params.threadId, format: 'full' }
        const { data } = await gmail.users.threads.get(threadParams)
        thread = data
      }
    
      const message = []
      if (params.to?.length) message.push(`To: ${wrapTextBody(params.to.join(', '))}`)
      if (params.cc?.length) message.push(`Cc: ${wrapTextBody(params.cc.join(', '))}`)
      if (params.bcc?.length) message.push(`Bcc: ${wrapTextBody(params.bcc.join(', '))}`)
      if (thread) {
        message.push(...getThreadHeaders(thread).map(header => wrapTextBody(header)))
      } else if (params.subject) {
        message.push(`Subject: ${wrapTextBody(params.subject)}`)
      } else {
        message.push('Subject: (No Subject)')
      }
      message.push('Content-Type: text/plain; charset="UTF-8"')
      message.push('Content-Transfer-Encoding: quoted-printable')
      message.push('MIME-Version: 1.0')
      message.push('')
    
      if (params.body) message.push(wrapTextBody(params.body))
    
      if (thread) {
        const quotedContent = getQuotedContent(thread)
        if (quotedContent) {
          message.push('')
          message.push(wrapTextBody(quotedContent))
        }
      }
    
      return Buffer.from(message.join('\r\n')).toString('base64url').replace(/\+/g, '-').replace(/\//g, '_').replace(/=+$/, '')
    }
  • Shared helper function used by all tools to handle OAuth2 client creation, credential validation, Gmail client setup, API call execution, and error handling.
    const handleTool = async (queryConfig: Record<string, any> | undefined, apiCall: (gmail: gmail_v1.Gmail) => Promise<any>) => {
      try {
        const oauth2Client = queryConfig ? createOAuth2Client(queryConfig) : defaultOAuth2Client
        if (!oauth2Client) throw new Error('OAuth2 client could not be created, please check your credentials')
    
        const credentialsAreValid = await validateCredentials(oauth2Client)
        if (!credentialsAreValid) throw new Error('OAuth2 credentials are invalid, please re-authenticate')
    
        const gmailClient = queryConfig ? google.gmail({ version: 'v1', auth: oauth2Client }) : defaultGmailClient
        if (!gmailClient) throw new Error('Gmail client could not be created, please check your credentials')
    
        const result = await apiCall(gmailClient)
        return result
      } catch (error: any) {
        return `Tool execution failed: ${error.message}`
      }
    }

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed2 schema fields changedv1.0.0
    • addedInput schema / $schema
      Added value: +"http://json-schema.org/draft-07/schema#"
    • addedInput schema / additionalProperties
      Added value: +false
  2. First observed

TDQS

B3.1/5.0
Behavior2/5

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

No annotations provided, so the description carries full burden. It mentions 'mechanics of the raw parameter' but fails to disclose that the raw parameter overrides other fields, which is a critical behavioral trait.

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

Conciseness4/5

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

The description is concise with two short sentences, front-loading the main purpose. No unnecessary information, though it could be slightly more structured.

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

Completeness2/5

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

Given no output schema and the complexity of the raw parameter overriding, the description is incomplete. It does not explain the implications of the raw parameter or provide a complete picture of the tool's behavior.

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?

Schema coverage is 100% with detailed descriptions for each parameter. The description adds minimal value by drawing attention to the raw parameter, but does not significantly enhance understanding beyond 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?

The description clearly states the action ('Create a draft email') and the resource ('in Gmail'), which is specific and distinguishable from sibling tools like 'send_draft' or 'delete_draft'.

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

Usage Guidelines2/5

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

No guidance on when to use this tool versus alternatives (e.g., send_draft). The description only hints at the raw parameter mechanics but does not provide usage context.

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