Kai Email System logo

Kai Email System

Community
cgallic
kai-email-system

Plan and batch-produce an entire email system for a product. Maps every lifecycle touchpoint (onboarding, activation, conversion, retention, transactional, win-back), generates all emails with quality gates, outputs Loops-ready copy. Use when "create all emails", "email system", "build email sequences", "lifecycle emails", "onboarding sequence", "set up transactional emails", "plan all the emails for [product]", or any request to systematically produce a complete set of emails for a platform.

Overview

Publishercgallic
Repositorykai-cmo-harness
Skill namekai-email-system
Stars
61
Forks
8
Bundled files
Instructions only
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by cgallic on GitHub. Read the source before you install it.

Installation

Install the Kai Email System AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/cgallic/kai-cmo-harness/tree/main/harness/skills-v2/kai-email-system
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/cgallic/kai-cmo-harness.git /tmp/kai-cmo-harness
mkdir -p .claude/skills
cp -r /tmp/kai-cmo-harness/harness/skills-v2/kai-email-system .claude/skills/kai-email-system
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Kai Email System in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Kai Email System on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Kai Email System is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Objective

A complete, gated email system for a product: an approved lifecycle map covering every touchpoint the product actually has, plus produced copy for each email in priority order, plus the Loops implementation notes an operator needs to wire the triggers. All emails target Loops (lifecycle + transactional). The map is the contract — the emails are what the map spawns.

Done when

Work type email-lifecycle — floor E5/C3/O3 (harness/eco-floors.yaml). Any cold outreach email inside the system is graded under cold-email instead (floor E5/C4/O3).

  • E5 — the ESP reports the send completed with a recipient count reconciled to the approved segment. This skill's local output reaches E3 at best: a named human approved the lifecycle map and the exact copy. The floor closes when the sends go out through Loops, which happens outside this skill.
  • C3 — every email passes its declared gates (below), and a named non-producer read the set end to end. Unsubscribe and sender-identity compliance is a C4 field-standard item rather than a lint — check it against harness/references/advertising-compliance.md.
  • O3 — open rate, click rate, reply rate, unsubscribe rate, and revenue per recipient read from the ESP connector at 14 days. Record the baseline and threshold before the first send, never after.

Constraints

  • MARKETING.md in the project root is the product bible. Read it before asking the user anything. If it does not exist, build it from the codebase and confirm the draft before saving — never open with "what is the product?". Discovery material: CLAUDE.md, README.md, PROJECT.md, package.json and any project description files in the project root and additional working directories; existing email templates, Loops config, or email code (grep for "loops", "resend", "email", "transactional"); landing pages and marketing copy for the value prop; user flows, auth systems, billing integrations, database schemas, API routes; existing content, social profiles, marketing materials.

  • MARKETING.md required sections (this file is the canonical template other /kai skills point to; unknowns stay [unknown — ask user]):

    markdown
    # MARKETING.md — Product Marketing Bible
    > Auto-generated by the marketing harness. Updated [date]. Edit freely — all /kai skills read this.
    
    ## Product
    Name · one-liner · category · URL
    
    ## ICP
    ### Primary ICP
    Title/role · company size · industry · budget authority · trigger event (what makes them look NOW)
    · current solution (including "nothing") · pain severity 1-10
    ### Secondary ICP (if applicable) — same fields
    ### Anti-ICP — who NOT to target, and why
    
    ## Personas
    Map to the closest harness personas in `knowledge/personas/_persona-index.md`, then customize.
    Per persona: who they are · core frustration · language they actually use (pull from reviews,
    forums, support tickets) · top 3 objections · decision trigger · channels they trust.
    ### User Segments (product-level) — e.g. admin who sets up the account, member who uses daily
    
    ## Value Prop
    Core transformation (before vs after) · 3-5 key actions · activation moment · aha moment
    
    ## Business Model
    Monetization and what triggers upgrade · pricing · stage (pre-launch/early/growth/scale) + MRR
    
    ## Current Marketing
    Channels active · email platform · ad platforms · social · content surfaces
    
    ## Brand Voice
    Tone · banned patterns · 1-2 example sentences in the right voice
    
    ## Competitive Landscape
    Top 3 competitors · our differentiation
    
    ## Key Metrics (if known)
    Traffic · conversion rate · email list size · MRR/ARR
  • The lifecycle map is approved before any email is written. Present it and know three things before producing: what is missing, which emails are definitely not needed, and which priority tier to produce first. Map output: workspace/email-system-map.md.

  • Every email in the map specifies email name (becomes the Loops template name), trigger (Loops event name, e.g. user.signed_up, trial.expiring_3d), stage, priority (P0 = launch blocker, P1 = week 1, P2 = month 1), and segment.

  • Load before writing: knowledge/channels/email-lifecycle.md, harness/skill-contracts/email-lifecycle.yaml, harness/skill-contracts/email.yaml. For cold outreach emails also harness/skill-contracts/cold-email.yaml and harness/references/cold-email-rules.md.

  • Per-email format contract — each email carries trigger, segment, stage, send timing, a primary subject plus two variants, preview text, body, one CTA, and its gate results:

    1. Four U's >= 10/16, scored 1-4 per U.
    2. Zero banned words (Tier 1: leverage, utilize, synergy, innovative, deep dive, circle back, touch base, moving forward, at the end of the day).
    3. Zero AI slop ("In conclusion", "It's important to note", "In today's rapidly evolving", "This comprehensive guide", "Without further ado", "It's worth noting that").
    4. Subject line <= 50 chars (mobile truncation).
    5. Preview text 40-90 chars.
    6. Body <= 500 words (200-500 target).
    7. Single primary CTA — one action per email.
    8. First line not generic (no "I hope this email finds you well").
  • Retry discipline: max 2 retries per email, fixing only the named failing dimension. After 2 failures, flag the email in the quality report and move on.

  • Produce in priority order, P0 first. Emails inside one sequence are written in order because tone has to flow; emails in different stages are independent and can be produced in parallel.

  • Persona application: if MARKETING.md names a persona, load the full profile from knowledge/personas/ and match tone to that persona's language patterns and pain points. Otherwise use the brand voice from MARKETING.md.

  • Nothing sends. This skill produces copy and setup notes; the send happens in Loops after human approval.

Context

NeedLoad
Lifecycle patterns, subject-line formulas, anti-patternsknowledge/channels/email-lifecycle.md
Gate thresholds and output format for lifecycle emailharness/skill-contracts/email-lifecycle.yaml
General email contractharness/skill-contracts/email.yaml
Cold outreach contract and lawharness/skill-contracts/cold-email.yaml + harness/references/cold-email-rules.md
Unsubscribe, sender identity, CAN-SPAM/GDPRharness/references/advertising-compliance.md
Persona profiles and language patternsknowledge/personas/_persona-index.md + the matching file in knowledge/personas/
Product, ICP, voice, current channelsMARKETING.md (project root)

Standard B2B SaaS stage map — adapt to the actual product, removing irrelevant stages and adding product-specific ones:

StageTrigger typeEmails
WelcomeTransactionalWelcome + getting started
OnboardingBehavioralSetup incomplete, first value moment, invite team
ActivationBehavioralFeature discovery, usage milestone
ConversionTime + behavioralTrial expiring (T-3, T-1, T-0), upgrade prompt
EngagementBehavioral + scheduledWeekly digest, new feature announcement
RetentionBehavioralUsage drop, re-engagement, churn risk
Win-backTime-basedChurned user sequence (D+7, D+14, D+30)
TransactionalEvent-drivenPassword reset, invoice, plan change, security alert
ReferralMilestoneReferral ask after value moment

Output. Emails go to workspace/emails/[stage]/[email-name].md — e.g. workspace/emails/onboarding/setup-incomplete.md. Alongside them:

  • workspace/emails/_email-system-map.md — the full lifecycle map.
  • workspace/emails/_quality-report.md — totals (emails produced, passed, flagged, average Four U's), a per-email row with each gate result, and every email that failed after 2 retries with its specific failure reason.
  • workspace/emails/_loops-setup.md — every trigger event to create in Loops; segments/groups needed; sequence flows and the delays between chained emails (onboarding, trial expiration, win-back); and which emails are transactional (no unsubscribe) versus marketing (unsubscribe required).

Escalate when

  • The product's activation moment or monetization trigger cannot be determined from the codebase, so behavioral triggers would be guesses.
  • An email would make a quantitative or compliance-bearing claim (pricing, guarantees, results) that MARKETING.md does not support.
  • The transactional/marketing split is ambiguous for an email — misclassifying it is an unsubscribe compliance failure.
  • Two retries failed on the same email for the same reason.
  • The user wants sends scheduled or triggers activated in Loops from here.

Frequently asked questions

What does the Kai Email System AI skill do?

Plan and batch-produce an entire email system for a product. Maps every lifecycle touchpoint (onboarding, activation, conversion, retention, transactional, win-back), generates all emails with quality gates, outputs Loops-ready copy. Use when "create all emails", "email system", "build email sequences", "lifecycle emails", "onboarding sequence", "set up transactional emails", "plan all the emails for [product]", or any request to systematically produce a complete set of emails for a platform.

Why use Kai Email System on TypingMind?

Because you install it once and use it with any model. Kai Email System is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Kai Email System in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/cgallic/kai-cmo-harness/tree/main/harness/skills-v2/kai-email-system. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Kai Email System?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Kai Email System?

As many as you like. As long as a model supports skills, you can use Kai Email System with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Kai Email System AI skill free?

It is published on GitHub by cgallic. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇