Generates email sequences for different contexts. Currently supports event follow-up sequences. Produces ready-to-send emails with subject lines, body copy, and timing guidance.
The English summary is being prepared. Showing the original description.
See how to installReview the instructions the agent will receive before you install.
This skill generates email sequences tailored to specific contexts. Each email type has its own rules derived from sequences that performed well in production.
Before generating any content, ask the user:
What type of email sequence do you need?
Wait for the user to select before proceeding. If they select a type marked "coming soon," tell them it's not yet available and ask if they want event follow-up instead.
First, read from the repo. Most of what you need is already documented:
| What you need | Where to find it |
|---|---|
| Product name | docs/inputs/product_brief.md |
| Product capabilities | docs/inputs/product_brief.md (key capabilities section) |
| Pain points | docs/inputs/messaging_positioning.md (value propositions, pain points) |
| Social proof / customers | docs/inputs/testimonials.md (approved quotes and customer names) |
| Target persona context | docs/inputs/target_personas.md |
| Asset links | docs/inputs/product_brief.md (links section) |
Read these files BEFORE asking the user anything. Extract:
Then ask only for what the repo cannot tell you:
If the user provides conversation-specific pain points, use those. Otherwise, use the pain points from the messaging doc.
Note: Sender first name and company name are filled dynamically by the user's email software. Use placeholders: [Sender First Name] and [Company Name].
Generate a 4-email sequence following this exact structure. Each email has ONE job and ONE CTA.
Timing: Send within 1 to 2 days of the event Thread: New thread Job: Remind them of the conversation and re-surface the pain
Subject line pattern:
Structure:
Hey [First Name],
At [event name + venue detail], I talked to a number of folks about [topic]. These kept coming up:
• [Pain point 1]
• [Pain point 2]
• [Pain point 3]
[Single direct question: "Is your team running into any of these?"]
- [Sender First Name]
Rules:
[Sender First Name] placeholder, preceded by a single dashTiming: +2 work days (if no reply) Thread: Reply to Email 1 Job: Add social proof from peers and introduce the product
Structure:
[First Name], other [role/title]s there were telling us:
"[Paraphrased quote about the problem]"
Why? Because [specific explanation]. [Company examples]. They all had one thing in common: [root cause].
[Product name] [solves this] with:
• [Capability 1]
• [Capability 2]
• [Capability 3]
[CTA: Link to demo video or short resource]
- [Sender First Name]
Rules:
Timing: +4 work days (if no reply) Thread: NEW thread (breaks the reply chain) Job: Lead with customer results, offer a resource
Subject line pattern:
Structure:
Hi [First Name],
[Customer names] have [achieved specific outcome].
We pulled their lessons into a short [playbook/guide]. It's practical, step-by-step guidance from [role]s who've already solved the problem.
[CTA: Link to guide]
- [Sender First Name]
P.S. Let me know if there's a specific challenge you're tackling, I can point you to the most relevant parts.
Rules:
Timing: +5 work days (if no reply) Thread: Reply to Email 1 (original thread) Job: Give permission to not respond, but leave door open
Structure:
Hi [First Name],
We don't force [their team type] to [do painful thing].
When you're ready to [achieve desired outcome], like the ones below, reach out. I'll be happy to show you how.
• [Problem/attack type 1]
• [Problem/attack type 2]
• [Problem/attack type 3]
• [Problem/attack type 4]
• [Problem/attack type 5]
• [Problem/attack type 6]
- [Sender First Name]
Rules:
These sequences are for programmatic outreach (one template, many recipients). Personalization must be real, not fabricated.
Do:
Don't:
docs/inputs/testimonials.md or docs/inputs/product_brief.md. If no approved customers exist, use generic framing ("[Company type]s like yours") or ask the user for approved names.The goal is to feel personal because it references real shared context, not because it pretends to remember a conversation that may not have happened.
All links must be real, working URLs. Never use placeholder text like "[link]", "(link)", or "link".
Where to get links:
docs/inputs/product_brief.md for approved asset URLs (demos, docs, guides)Link distribution across the sequence:
| CTA Type | Preferred Link | |
|---|---|---|
| 1 | No link | Question only |
| 2 | Demo video | Product demo or explainer video |
| 3 | Resource | Guide, playbook, or beyondidentity.ai landing page |
| 4 | No link | "Reach out" only |
beyondidentity.ai usage:
Never:
Always write the final sequence to a plain text file in output/email/. This is the primary deliverable. The file must be copy/paste ready for Google Docs with no Markdown formatting artifacts. Only promote to docs/outputs/ if the user explicitly requests it.
File naming: email-sequence-[event-or-context]-[product].txt (e.g., email-sequence-rsac-ceros.txt)
File format rules:
.txt file, no Markdown syntax (no ##, **, ---, or []() links)EMAIL 1: THE RECALL)Send timing:, Thread:, Subject:)- prefixFor each email in the file, use this structure:
EMAIL [N]: [NAME]
Send timing: [When to send]
Thread: [New thread / Reply to Email X]
Subject: [Subject line or "N/A (threaded reply)"]
[Full email body]
After all 4 emails, include a plain text summary:
SEQUENCE SUMMARY
Email 1 | Day 0 | New thread | [CTA type]
Email 2 | +2 days | Reply | [CTA type]
Email 3 | +6 days | New thread | [CTA type]
Email 4 | +11 days | Reply | [CTA type]
In addition to the file, display the sequence in the conversation using Markdown for readability. But the file in output/email/ is the canonical output.
Before delivering, verify each email passes:
When docs are incomplete:
product_brief.md is empty or a template, ask the user for product name and capabilitiestestimonials.md has no approved quotes, ask the user for customer names or use "[Company type]s like yours"messaging_positioning.md is empty, ask the user for pain pointsSequence adjustments:
No reviews yet. Tell others what you thought.
No login needed. Share how it worked for you, ask a question, or note where it did not run.
No comments yet. Be the first to write one.
Generates paid ad copy for Google, Meta, and LinkedIn. Produces platform-ready variants using a creative matrix approach with headlines, descriptions, and image briefs.
The English summary is being prepared. Showing the original description.
Use when you want paid ad copy for Google, Meta, or LinkedIn, including headlines, descriptions, and image briefs.
The English summary is being prepared. Showing the original description.
Use when you want campaign performance reviewed against benchmarks, with a health score and prioritized optimization recommendations.
The English summary is being prepared. Showing the original description.
Use when you want a strict review of a marketing asset against style rules, factual accuracy, and messaging alignment.
The English summary is being prepared. Showing the original description.
Skill Optimization: Auto Research. Autonomously optimize any Claude Code skill by running it repeatedly, scoring outputs against binary evals, mutating the prompt, and keeping improvements. Based on Karpathy's autoresearch methodology. Use when: optimize this skill, improve this skill, run autoresearch on, make this skill better, self-improve skill, benchmark skill, eval my skill, run evals on. Outputs: an improved SKILL.md, a results log, and a changelog of every mutation tried.
The English summary is being prepared. Showing the original description.
Use when you want to benchmark and improve an existing repo skill through repeated evals, prompt mutations, and score tracking.
The English summary is being prepared. Showing the original description.