Skip to main content
Glama

Create automation

create_automation
Destructive

Create a triggered text flow (as a draft): when trigger.event is tracked, run steps such as send_sms, wait, wait_for_event, and condition branches. Example: on cart_abandoned, wait 3600s, then if checkout_completed was not received send a reminder. A send_sms step with send_time: "optimal_user_time" and max_wait_seconds (60 to 86400, required) waits for the hour that person usually replies or clicks, 10am their time with no history, inside quiet hours: recommend it for marketing and lifecycle texts, leave it off anything time-sensitive, and it is refused for one-time codes and order or shipping texts. For events that repeat, trigger.cooldown_seconds stops the same person starting it again for that long. Merge fields like {{plan}} fill from the trigger event's properties; issues names any no tracked event carries. Leave out from and TextKit picks a number approved for category (marketing, transactional for alerts, otp for codes) when it sends; sender shows the numbers, and issues says so when none is approved yet. Pass from only to pin a number. Drafts send nothing until activated.

Tool call template

This tool declares no required parameters. Add optional parameters from the input schema as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {},
    "name": "create_automation"
  }
}

Input Schema

NameRequiredDescriptionDefault
fromNoOne of your numbers, to pin the flow to it. Leave it out and TextKit picks a number approved for `category`. A sandbox number makes it a sandbox flow.
nameNoAutomation name
stepsNoUp to 20 steps, run in order
triggerNoWhat starts a run: an event, with optional filters
categoryNoWhat it sends: marketing, alerts (transactional) or codes (otp). With no `from`, TextKit picks a number approved for this. Falls back to a built-in topic.
topic_idNoSkip recipients unsubscribed from this topic.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed3 schema fields changed
    • changedInput schema / properties / steps / items / description
      Previous value: -"One of: send_sms {body, from?}; wait {seconds}; wait_for_event {event, timeout_seconds, on_received, on_timeout}; condition {check, on_met, on_not_met}; end. Branches are \"continue\", \"end\", or \"goto:<id of a later step>\"."New value: +"One of: send_sms {body, from?, send_time?, max_wait_seconds?}; wait {seconds}; wait_for_event {event, timeout_seconds, on_received, on_timeout}; condition {check, on_met, on_not_met}; end. Branches are \"continue\", \"end\", or \"goto:<id of a later step>\"."
    • addedInput schema / properties / steps / items / properties / max_wait_seconds
      Added value: +{
      +  "description": "send_sms with optimal_user_time, required: the longest this text may wait for that hour",
      +  "maximum": 86400,
      +  "minimum": 60,
      +  "type": "integer"
      +}
    • addedInput schema / properties / steps / items / properties / send_time
      Added value: +{
      +  "description": "send_sms: `optimal_user_time` waits for the hour this person usually replies or clicks, inside quiet hours. Default: right_away. Refused for one-time codes and order or shipping texts.",
      +  "enum": [
      +    "right_away",
      +    "optimal_user_time"
      +  ],
      +  "type": "string"
      +}
  2. Changed2 schema fields changed
    • addedInput schema / properties / category
      Added value: +{
      +  "description": "What it sends: marketing, alerts (transactional) or codes (otp). With no `from`, TextKit picks a number approved for this. Falls back to a built-in topic.",
      +  "enum": [
      +    "marketing",
      +    "transactional",
      +    "otp"
      +  ],
      +  "nullable": true,
      +  "type": "string"
      +}
    • changedInput schema / properties / from / description
      Previous value: -"One of your numbers. A sandbox number makes it a sandbox flow."New value: +"One of your numbers, to pin the flow to it. Leave it out and TextKit picks a number approved for `category`. A sandbox number makes it a sandbox flow."
  3. Changed16 schema fields changed
    • addedInput schema / properties / name / description
      Added value: +"Automation name"
    • addedInput schema / properties / steps / description
      Added value: +"Up to 20 steps, run in order"
    • addedInput schema / properties / steps / items / properties / event / description
      Added value: +"wait_for_event: the event to wait for"
    • addedInput schema / properties / steps / items / properties / on_met / description
      Added value: +"condition: where to go when the check passes: continue, end, or goto:<step id>"
    • addedInput schema / properties / steps / items / properties / on_not_met / description
      Added value: +"condition: where to go when the check fails: continue, end, or goto:<step id>"
    • addedInput schema / properties / steps / items / properties / on_received / description
      Added value: +"wait_for_event: where to go when the event arrives: continue, end, or goto:<step id>"
    • addedInput schema / properties / steps / items / properties / on_timeout / description
      Added value: +"wait_for_event: where to go when time runs out: continue, end, or goto:<step id>"
    • addedInput schema / properties / steps / items / properties / seconds / description
      Added value: +"wait: how long, from 60 seconds to 30 days"
    • addedInput schema / properties / steps / items / properties / timeout_seconds / description
      Added value: +"wait_for_event: how long to wait, from 60 seconds to 30 days"
    • addedInput schema / properties / steps / items / properties / type / description
      Added value: +"What the step does"
    • addedInput schema / properties / trigger / description
      Added value: +"What starts a run: an event, with optional filters"
    • addedInput schema / properties / trigger / properties / event / description
      Added value: +"The tracked event that starts a run"
    • addedInput schema / properties / trigger / properties / filters / description
      Added value: +"Up to 5 checks on the event's properties. All must hold"
    • addedInput schema / properties / trigger / properties / filters / items / properties / op / description
      Added value: +"eq, neq, contains, gt, lt, exists or not_exists"
    • addedInput schema / properties / trigger / properties / filters / items / properties / property / description
      Added value: +"The event property to check"
    • addedInput schema / properties / trigger / properties / filters / items / properties / value / description
      Added value: +"The value to compare with. Not used by exists or not_exists"
  4. Changed1 schema field changed
    • addedInput schema / properties / trigger / properties / cooldown_seconds
      Added value: +{
      +  "description": "Optional. After a run starts for a person, they do not start another for this long (for events that repeat, like a paywall view).",
      +  "maximum": 7776000,
      +  "minimum": 60,
      +  "type": "integer"
      +}
  5. First observed

TDQS

A4.4/5.0
Behavior5/5

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

Annotations declare destructiveHint=true and non-idempotent, but the description adds crucial mitigating context: the flow is created as a draft that sends nothing until activated. It also discloses refusal rules for send_time, automatic sender selection by category, cooldown deduplication, and that `issues`/`sender` report problems and approved numbers. That is substantive behavior beyond the structured fields.

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

Conciseness3/5

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

Purpose is front-loaded in the first clause, but the body is a dense run-on paragraph that repeats schema details (max_wait_seconds range, send_time refusal rules) and chains many ideas with commas and semicolons. It is information-rich but harder to scan than it needs to be for a tool of this complexity.

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 complex nested tool with no output schema, the description covers the draft lifecycle, sender/category resolution, merge-field resolution, cooldown semantics, and the existence of an `issues` diagnostics channel. It does not state what identifier is returned on creation or how validation failures surface as errors, which is the main remaining gap.

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; the description earns above that by explaining cross-parameter behavior the schema only states piecemeal: merge fields resolve from the trigger event's properties, leaving out `from` triggers category-based number selection, and max_wait_seconds is required with optimal_user_time. Some of this repeats schema text, but the interaction between category, from, and send_time is genuinely added meaning.

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 ('Create a triggered text flow (as a draft)') and immediately enumerates the step types (send_sms, wait, wait_for_event, condition) that define the artifact. An agent can distinguish this from update_automation, preview_automation, or test_automation without opening any 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 conditional guidance: recommend optimal_user_time for marketing/lifecycle texts, leave it off time-sensitive sends, and it is refused for OTP and order/shipping texts. Also clarifies drafts send nothing until activated, implying activate_automation as the follow-up. It stops short of explicitly naming sibling tools like preview_automation or test_automation as validation steps.

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

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.