Skip to main content
Glama
vcnngr

gwagent

by vcnngr
OverviewSchemaSourceRelated ServersDiscussionsAdmin
Python
Local

gmail_send

Send new Gmail emails with optional Cc, Bcc, attachments, From alias, or scheduled delivery. Recipients are verified before sending; immediate undo can recall a message.

Instructions

Send a new e-mail (any Unicode text) with optional Cc, Bcc, attachments, From alias, or schedule it.

The account's signature is added after the body by Gmail (signature=false removes it). Recipients are verified on the compose screen before sending; on any mismatch nothing is sent. Scheduled mail waits in the 'Scheduled' folder (gmail.scheduled, gmail.cancel_scheduled).

Send is never repeated blind: when Gmail's reaction to Send is unclear the mail is looked up in Outbox / Sent; if that cannot tell, the op fails with error 'send_state_unknown' and nothing is re-sent (check gmail.list folder='sent' before any retry). A result may carry verified_in='outbox' | 'sent' when the send was confirmed that way; a scheduled send counts only once Gmail confirms it or it is found in 'Scheduled' (verified_in='scheduled').

Gmail offers 'Undo' for about 7 s after 'Sent' (undo_available tells whether it was still on screen when the op returned; gmail.undo uses it). undo=true taps that Undo within this op, the moment Gmail shows it: the result then has undone='send' and saved_as_draft=true only when the mail came back as a draft and is not in Outbox / Sent; undone=false (with undo_error) means it was sent anyway.

Requires confirm=true (outward-facing or destructive).

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
ccNoCc addresses.
toYesRecipient addresses.
bccNoBcc addresses.
bodyNoPlain-text body (newlines allowed).
undoNoRecall the mail with Gmail's 'Undo' right after sending (it comes back as a draft and is not delivered); not with schedule.
confirmNoMust be true: this action is outward-facing or destructive.
subjectNoSubject line.
scheduleNoSend later instead of now: 'YYYY-MM-DD HH:MM' (24h, the phone's time zone) or a preset offered by Gmail's Schedule send dialog ('tomorrow morning', 'tomorrow afternoon', 'monday morning', ...).
signatureNoKeep the signature Gmail appends after the body.
attachmentsNoPaths of files on this computer to attach (pushed to the phone first).
drive_filesNoGoogle Drive files to insert as links: 'name' (found with Drive search) or '/Folder/name' (from My Drive). If a recipient has no access Gmail asks about sharing: the mail is then kept as a draft and an error is returned (share it with drive.share, then gmail.send_draft).
from_addressNoSend from this address (an alias / delegated address of the account).

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observedv0.1.0

TDQS

A4.6/5.0
Behavior5/5

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

With no annotations, the description carries the full behavioral burden and does so thoroughly: it discloses recipient verification, signature handling, send-confirmation logic, the 'send_state_unknown' failure mode, verified_in values, Gmail's ~7s Undo window, undo result fields, and confirm=true requirement. No annotation contradiction exists.

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 long but front-loads the core action and then layers behavioral details that are relevant for a 12-parameter, outward-facing tool. It could be tightened or structured with clearer headings, but every sentence conveys useful operational context.

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

Completeness5/5

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

Given the tool's complexity (12 parameters, no annotations, no output schema), the description is complete enough: it covers prerequisites, send verification, error states, undo semantics, scheduling, signature behavior, and drive-file sharing edge cases. An agent has what it needs to invoke the tool correctly and handle failure modes.

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

Parameters4/5

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

Schema coverage is 100%, so the baseline is 3, but the description adds meaningful semantics beyond the schema for undo (not with schedule), schedule presets/timezone, signature, attachments, drive_files sharing failures, and from_address aliases. It does not add much for to/cc/bcc/body/subject, but the added context is valuable.

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 ('Send a new e-mail') and distinguishes itself from reply/draft siblings by specifying new mail with optional Cc, Bcc, attachments, From alias, or scheduling. An agent can immediately tell what this tool does and how it differs from gmail_reply, gmail_draft, or gmail_send_draft.

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?

Provides clear context and prerequisites: requires confirm=true, describes scheduled-mail behavior and its relationship to gmail.scheduled/gmail.cancel_scheduled, and routes drive-sharing failures to drive.share + gmail.send_draft. It does not explicitly state when to prefer this over gmail_send_draft or gmail_reply, but the 'new e-mail' framing and named alternatives provide strong guidance.

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

Deploy Server

Other Tools