Invoke an ad-hoc automation with inline steps. Valid step actions: send, send-list, delay, cancel, update-profile, invoke, fetch-data. To cancel a previously started automation, use the cancel_automation tool instead.
| Name | Required | Description | Default |
|---|---|---|---|
| data | No | ||
| brand | No | ||
| profile | No | ||
| template | No | ||
| recipient | No | ||
| automation | Yes | The automation definition with typed steps |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already say this is not read-only and not idempotent, but the description adds no behavioral context about what invoking does (e.g., executes steps immediately, sends messages, updates profiles) or what happens to previously started automations. The cancel guidance contradicts the presence of a cancel step, further muddying behavior.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is compact and front-loaded with the core purpose, and the action list is a useful summary. The cancel sentence is short but creates ambiguity, which slightly reduces the value of the concise structure.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a tool with a nested automation definition, seven step variants, no output schema, and sparse parameter descriptions, the description is too thin. It omits return/execution behavior, token semantics, and parameter meaning, so an agent would struggle to construct a correct invocation despite the detailed schema.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is only 17%, so the description carries the burden for explaining parameters, but it only names the step actions. Top-level parameters like data, brand, profile, template, and recipient are left undocumented in the description; the nested schema has to do all the work.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description uses a specific verb ('Invoke') and resource ('ad-hoc automation') plus 'with inline steps', which distinguishes it from sibling invoke_automation_template. It also enumerates the valid step actions, making the tool's scope immediately recognizable.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
It explicitly routes cancellation to cancel_automation, which is good alternative guidance. However, it also lists 'cancel' as a valid inline step action and the schema's cancel step accepts a cancelation_token, creating conflicting guidance about when to use this tool versus the sibling. No guidance is given for choosing between ad-hoc inline automations and invoke_automation_template.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Add one secure layer between your agents and this server.
Tools are mostly organized as distinct resource/action pairs, but several clusters are easy to confuse: list subscription tools (add_subscribers_to_list vs bulk_subscribe_to_list vs subscribe_user_to_list), message vs message-content vs message-history retrieval, and the many journey/journey-template list/get tools. Detailed descriptions rescue most selections, but the sheer number of near-identical verb/resource names creates real misselection risk.
Almost all tools follow a snake_case verb_noun pattern (create_, get_, list_, replace_, send_, publish_, archive_). Minor deviations keep it from a perfect score: courier_installation_guide is noun-first, and add_bulk_users sits awkwardly next to the bulk_add_* family, but the overall convention is predictable and readable.
144 tools is an extreme working-set size for an agent to hold and choose from, far beyond the reasonable 3–15 range. Even for a broad platform like Courier, this should be split into focused sub-servers (templates, journeys, users, lists, preferences, etc.) to remain usable.
The surface is remarkably comprehensive, covering sending, templates, journeys, automations, users, tenants, lists, preferences, providers, routing, brands, audiences, translations, digests, bulk jobs, and audit events. Notable gaps exist—automation template CRUD and digest schedule management are missing—but most workflows can still be completed with workarounds.
