Skip to main content
Glama
agentisend

agentisend

Preflight an email

preflight_email
Idempotent

Dry-run an email before sending: validate every gate—domain, suppression, budget, loop guard, content scan—and preview the exact ok/error response without spending budget or delivering anything.

Instructions

Purpose: find out exactly what a send would do, without doing it. Runs every gate a real send runs — validation, sending domain, sandbox, trust, suppression, budget, loop guard, content scan. NOT for: sending. Nothing leaves and no budget is spent, ever. Free and unlimited on purpose. Returns: the same ok/error POST /emails would return for the same payload, plus each gate's own verdict, and simulated: true when every recipient is a simulator address. The arguments are the same ones send_email takes: copies, headers, attachments, an inline image content_id, a template and a topic.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "to": "<to>"
    },
    "name": "preflight_email"
  }
}

Input Schema

NameRequiredDescriptionDefault
ccNoCarbon-copy addresses — one string or a list.
toYesOne recipient address, or a list of up to 50. A bare string is fine.
bccNoBlind carbon-copy addresses — one string or a list.
fromNoSender address, e.g. "Acme <hello@acme.com>". Optional when template_id names a template with a default sender; one given here overrides it.
htmlNoHTML body. Provide html, text or template_id.
tagsNoYour own labels, returned on every event for this message.
textNoPlain-text body. Provide html, text or template_id.
headersNoExtra RFC 5322 headers.
subjectNoSubject line. Optional when template_id names a template that has one.
reply_toNoWhere replies should go, if not the from address.
timezoneNoTime zone for a scheduled_at phrase such as "tomorrow 9am". An ISO time ignores this. Example: America/Los_Angeles. Omitted phrases are read in UTC.
topic_idNoSubscription topic. Naming one screens every recipient against what they said about that topic.
attachmentsNoFiles to attach (max 100). One with content_id is an inline image.
template_idNoThe id or the alias of a published template to send instead of an inline body. Its subject, from and reply_to are used unless you pass your own.
scheduled_atNoISO 8601 time, or a short phrase ("in 1 hour", "tomorrow 9am").
idempotency_keyNoYour own key for this action. Retrying with the same key returns the first result instead of doing it twice — send one whenever a retry could double the effect.
template_valuesNoValues for the template variables.

Output Schema

NameRequiredDescriptionDefault
okNoWhat the real send would answer.
errorNoThe refusal the real send would return, or null.
checksNoEach gate by name, with what it found.
simulatedNoTrue when every recipient is a simulator address.

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observedv0.1.0

TDQS

A4.6/5.0
Behavior4/5

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

Beyond the annotations, the description discloses that every gate in a real send runs (validation, domain, sandbox, trust, suppression, budget, loop guard, content scan), that no budget is consumed, and that the tool is free, unlimited, and safe to retry. It also explains the simulated: true flag. Residual tension: readOnlyHint=false understates that nothing is mutated, though this is a mild mismatch, not a contradiction; the description still omits any auth or latency expectations.

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 terse and well structured around Purpose / NOT for / Returns, with no filler sentences. The closing Returns sentence partly restates what the output schema already carries, so it is slightly less lean than it could be.

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?

For a 17-parameter tool, the description supplies everything an agent needs in order to call it correctly: the purpose, the safety and cost profile, the payload parity with send_email, and the shape of the result, which an output schema backs up. Nothing essential is missing.

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% across 17 parameters, so the baseline is 3, but the description earns an extra point by telling the agent that the arguments are identical to send_email's, letting existing knowledge transfer to this payload. It names only a sample of the fields (copies, headers, attachments, content_id, template, topic) rather than explaining the semantics of each.

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 gives a specific verb and resource ('find out exactly what a send would do, without doing it') and immediately distinguishes the tool from send_email by naming the full set of gates that it exercises without actually sending. An agent can tell what this is from the first sentence alone.

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

Usage Guidelines5/5

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

It states both sides of the routing decision: 'Purpose: ... without doing it' plus an explicit 'NOT for: sending', with send_email named as the tool whose payload semantics are mirrored. The exclusions ('nothing leaves, no budget is spent') make the when-to-use condition unambiguous.

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