Skip to main content
Glama

Add a step to a campaign

add_campaign_step
Destructive

Use this when the user wants another step in a campaign's sequence: a message on one channel, a LinkedIn profile visit or like, or a condition or A/B split that branches the sequence. Place it with sequenceNumber, in the main sequence or in a condition's lane. A manual step sends its subject and body as written, with {Variable} tokens filled in for each person; an AI step writes each message itself. On a running campaign, the step's messages go out to each person who reaches it. Once leads have moved through a branched sequence its shape cannot change. Writing an AI step can take a minute. Free. Calling again adds a second step.

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": {
      "campaignId": "<campaignId>",
      "channel": "<channel>",
      "sequenceNumber": "<sequenceNumber>",
      "daysBetween": "<daysBetween>"
    },
    "name": "add_campaign_step"
  }
}

Input Schema

NameRequiredDescriptionDefault
ccNoUp to 10 addresses copied on every email of the step.
bodyNoA `manual` step's message, up to 20,000 characters; the campaign's variable tokens, such as `{Contact First Name}`, are filled in for each lead. On a `linkedin_connection` step, the invitation note, up to 200 characters. Not on visits, likes and conditions.
typeNo`manual` (the default) sends `subject` and `body` as written; `ai_generated` writes each message for each lead from a topic it writes itself; `ai_reply` drafts a reply from `topic` that waits for your approval. A condition is always `manual`.
topicNoWhat the step should say, up to 500 characters. Needed by `ai_reply`; `ai_generated` writes its own.
channelYes`email`, `linkedin_connection` (an invitation), `linkedin_message` (to a connection), `linkedin_visit` (a profile visit), `linkedin_like` (a like on a recent post), `inmail`, or `condition`: a branch that waits for `conditionType` and sends each lead down its `yes` or `no` lane, or for an `ab_split`, down variant `a` or `b` at once. A `linkedin_connection` step cannot follow a `linkedin_message` step.
subjectNoA `manual` step's subject line, up to 500 characters. Not on visits, likes and conditions.
branchKeyNoThe lane of `parentStepId`: `yes` or `no` under a condition, `a` or `b` under an A/B split.
splitArmsNo`ab_split` only: both variants, `a` then `b`, each `{ key, label?, weight }`. 50 each when left out.
campaignIdYesThe campaign's id.
daysBetweenYesDays to wait after the step before (or after the start, for the first step), 0 to 90. On a condition, how long it waits for its signal, 1 to 30 days; an `ab_split` waits for nothing: send 0.
parentStepIdNoThe condition whose lane the step goes in, with `branchKey`. Left out, the step goes in the main sequence.
conditionTypeNoConditions only: `invite_accepted` waits for the invitation of the nearest `linkedin_connection` step before it, `clicked_link` for a click in the nearest email before it, and `ab_split` splits leads between variants `a` and `b` by `splitArms`. Conditions can be added only before the campaign starts.
sequenceNumberYesThe step's place, from 0, in the main sequence or in the lane `parentStepId` and `branchKey` name. Send the length of the sequence or lane to add at its end; a taken place answers `409` `step_position_taken`.

Schema Changelog

Changes observed during successful MCP inspections.

No schema history has been recorded yet.

TDQS

A4.3/5.0
Behavior5/5

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

Beyond the annotations (destructive, non-idempotent, open-world), the description discloses what actually happens: manual steps send subject/body with {Variable} tokens filled per person, AI steps write each message themselves, running campaigns dispatch to whoever reaches the step, AI generation can take a minute, the call is free, and calling again adds a second step rather than editing. These are exactly the operational facts an agent needs before invoking a mutating tool.

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 when-to-use condition and placement rule are front-loaded, and each sentence carries distinct information (step kinds, placement, manual vs AI, live-campaign behavior, branch lock, latency, cost, non-idempotency). The trailing fragments 'Free. Calling again adds a second step.' read tersely but still earn their place.

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 13-parameter destructive mutation with no output schema, the description covers the important behavioral surface: placement, live-campaign effects, irreversibility of branch shape, latency and cost. It does not mention authentication or permission requirements, and error behavior is only covered in the schema (409 step_position_taken), leaving a small gap.

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?

Schema coverage is 100% and the input schema already documents every parameter in depth, including channel enums, branchKey lanes, splitArms and daysBetween semantics. The description restates the manual-vs-AI body/topic behavior and the sequenceNumber placement rule, adding emphasis but little meaning not already in the schema, so the baseline 3 applies.

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 opens with a specific verb and resource ('another step in a campaign's sequence') and immediately enumerates the kinds of step it can be: a channel message, a profile visit or like, or a branching condition/A/B split. That level of specificity lets an agent separate this from update_campaign_step or delete_campaign_step without opening their schemas.

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?

It gives a clear when-to-use trigger ('when the user wants another step') and states real preconditions: conditions can only be added before the campaign starts, and a branched sequence's shape is locked once leads have moved through it. It does not explicitly name the update/delete siblings as the alternative, so it falls short of a 5, but the exclusions are concrete.

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.

Resources