Skip to main content
Glama

send_draft

Send a saved email draft from Gmail by specifying its ID to complete and deliver composed messages.

Instructions

Send an existing draft

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
idYesThe ID of the draft to send

Implementation Reference

  • src/index.ts:371-386 (registration)
    Registers the 'send_draft' MCP tool, including input schema (draft ID) and handler that calls Gmail API to send the draft.
    server.tool("send_draft",
      "Send an existing draft",
      {
        id: z.string().describe("The ID of the draft to send")
      },
      async (params) => {
        return handleTool(config, async (gmail: gmail_v1.Gmail) => {
          try {
            const { data } = await gmail.users.drafts.send({ userId: 'me', requestBody: { id: params.id } })
            return formatResponse(data)
          } catch (error) {
            return formatResponse({ error: 'Error sending draft, are you sure you have at least one recipient?' })
          }
        })
      }
    )
  • Inline handler function for 'send_draft' tool that uses handleTool to authenticate and execute gmail.users.drafts.send API call.
    async (params) => {
      return handleTool(config, async (gmail: gmail_v1.Gmail) => {
        try {
          const { data } = await gmail.users.drafts.send({ userId: 'me', requestBody: { id: params.id } })
          return formatResponse(data)
        } catch (error) {
          return formatResponse({ error: 'Error sending draft, are you sure you have at least one recipient?' })
        }
      })
    }
  • Zod schema defining input parameters for 'send_draft' tool: requires 'id' as string (draft ID).
    {
      id: z.string().describe("The ID of the draft to send")
    },
  • Shared helper function 'handleTool' used by 'send_draft' (and other tools) to handle OAuth2 authentication, credential validation, Gmail client creation, and API execution with 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) {
        // Check for specific authentication errors
        if (
          error.message?.includes("invalid_grant") ||
          error.message?.includes("refresh_token") ||
          error.message?.includes("invalid_client") ||
          error.message?.includes("unauthorized_client") ||
          error.code === 401 ||
          error.code === 403
        ) {
          return formatResponse({
            error: `Authentication failed: ${error.message}. Please re-authenticate by running: npx @shinzolabs/gmail-mcp auth`,
          });
        }
    
        return formatResponse({ error: `Tool execution failed: ${error.message}` });
      }
    }

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observed

TDQS

C2.9/5.0
Behavior2/5

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

No annotations are provided, and the description fails to disclose behavioral traits such as side effects (e.g., draft deletion after sending), authorization requirements, or error conditions. The agent has minimal insight beyond the basic action.

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 a single concise sentence with no waste. However, it lacks structure (e.g., no separation of purpose and usage).

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

Completeness3/5

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

Given the tool's simplicity (one parameter, no output schema), the description provides minimal context. It omits information about what happens after sending (e.g., draft status change) and potential failure modes, leaving gaps for an agent.

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?

The schema covers 100% of parameters with descriptions, so baseline is 3. The description adds no extra meaning beyond what the schema already provides for the 'id' parameter.

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

Purpose4/5

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

The description clearly states the action 'Send' and resource 'existing draft', distinguishing from siblings like create_draft and delete_draft. However, it does not explicitly differentiate from send_message, which could cause confusion.

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 is provided on when to use send_draft versus alternatives like send_message or create_draft. There are no prerequisites or exclusions mentioned, leaving the agent to infer usage.

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