Cognotik | EmailCampaignTask

⚙️ ExecutionConfig.json


{
  "campaign_goal": "Re-engage inactive users",
  "subject_matter": "New AI-powered bento grids",
  "num_emails": 3,
  "campaign_type": "re_engagement",
  "brand_voice": "friendly",
  "generate_subject_variants": true
}

→

👁️ Session UI (Complete Campaign)

Email 1 - Day 0

Subject Options (A/B Test):

[A] We've missed you! See what's new.
[B] Your dashboard just got an upgrade.

Hi {{first_name}},

It's been a while since you logged into Cognotik. We've been busy building the "Industrial" design system...

Explore the New UI

Live Results Showcase

Explore actual email campaign artifacts and markdown reports generated by this task.

Configuration Parameters

Field	Type	Description
`campaign_goal`*	String	The primary objective (e.g., "Launch new API features").
`subject_matter`*	String	The product, service, or topic being promoted.
`target_audience`	String	Description of demographics, roles, or pain points.
`num_emails`	Int	Number of emails in sequence (1-10). Default: `3`.
`campaign_type`	String	`welcome_series`, `nurture`, `sales`, `re_engagement`, `newsletter`.
`primary_cta`	String	The main action (e.g., `schedule_demo`, `make_purchase`).
`brand_voice`	String	Tone (e.g., `professional`, `friendly`, `authoritative`).
`body_length`	String	`short` (<150 words), `medium`, or `long`.
`send_intervals`	List<Int>	Days between emails (e.g., `[1, 3, 7]`).
`revision_passes`	Int	Number of automated quality improvement passes (0-5).
`input_files`	List<String>	File patterns (e.g. `*/.md`) to use as brand context.
`generate_subject_variants`	Boolean	Whether to generate multiple A/B test options.

Dependencies

None. Operates as a standalone generator or can ingest input_files for brand context.

Token Usage

High Consumption

Uses multiple LLM calls: Strategy (1), Outlines (1 per email), Generation (1 per email), Revisions (N per email).

Task Execution Lifecycle

Phase 1: Strategy Development

Analyzes audience and subject matter to create a progression logic and key messaging map.

Phase 2: Sequence Outlining

Plans the specific purpose, emotional tone, and CTA for every email in the sequence.

Phase 3: Content Generation

Writes the copy, generates A/B subject variants, and injects personalization tokens.

Phase 4: Revision (Optional)

Performs N passes of editing for clarity, conciseness, and brand consistency.

Embedded Invocation (Headless)

Use the UnifiedHarness to run this task programmatically in CI/CD or CLI tools.


val harness = UnifiedHarness(serverless = true)
harness.start()

val executionConfig = EmailCampaignTaskExecutionConfigData(
    campaign_goal = "Onboard new developers",
    subject_matter = "Cognotik SDK",
    num_emails = 5,
    brand_voice = "technical",
    revision_passes = 2
)

harness.runTask(
    taskType = EmailCampaignTask.EmailCampaign,
    executionConfig = executionConfig,
    workspace = File("./my-project")
)

Direct Instantiation


val task = EmailCampaignTask(
    orchestrationConfig = config,
    planTask = EmailCampaignTaskExecutionConfigData(
        campaign_goal = "Re-engage churned users",
        subject_matter = "New Feature Launch",
        campaign_type = "re_engagement"
    )
)

Test Case Example

Example of a validated configuration for testing:


{ "campaign_goal": "Test Goal", "subject_matter": "Test Subject", "num_emails": 1 }

Prompt Segment

The following logic is injected into the LLM context:


EmailCampaign - Generate multi-email marketing or outreach sequences.
- campaign_goal: Primary objective.
- subject_matter: Product or topic.
- target_audience: Who is receiving the emails.
- campaign_type: welcome_series, nurture, sales, etc.
- num_emails: Length of sequence (1-10).
- brand_voice: professional, friendly, etc.
- primary_cta: Main action desired.
- related_files: Brand guidelines or context.