email-drafting
Draft a professional, audience-aware email or reply with calibrated tone and a clear call to action. Use when the user needs one message, response, follow-up, support note, or meeting request; use sales-email-sequences for a coordinated multi-touch outbound campaign.
git clone --depth 1 https://github.com/seb1n/awesome-ai-agent-skills /tmp/email-drafting && cp -r /tmp/email-drafting/communication/email-drafting ~/.claude/skills/email-draftingSKILL.md
# Email Drafting
This skill enables an AI agent to compose professional, context-appropriate emails across a variety of scenarios. The agent analyzes the purpose, audience, and tone requirements, then produces a complete email with subject line, body, and call to action. It supports common email types including cold outreach, follow-up sequences, customer support responses, internal status updates, and meeting requests.
## Workflow
1. **Identify the email type and objective.** Determine the category of the email — cold outreach, follow-up, customer support reply, internal update, or meeting request. Each type has different structural expectations. A cold outreach email prioritizes a compelling hook, while a support response leads with empathy and resolution. Clarify the single desired outcome: a scheduled call, an acknowledged resolution, an informed team, etc.
2. **Profile the recipient and calibrate tone.** Gather details about the recipient: name, title, company, relationship history, and any previous interactions. Use this context to select the appropriate tone — formal for executive outreach, conversational for peer updates, empathetic for complaint responses, or urgent when time-sensitive action is required. Apply personalization tokens such as `{{first_name}}`, `{{company}}`, and `{{pain_point}}` when drafting templates.
3. **Compose the subject line.** Write a subject line that is specific, concise (under 60 characters), and action-oriented. Avoid generic phrases like "Quick question" or "Touching base." Instead, reference a concrete benefit or context point, e.g., "Cut your onboarding time by 40% — here's how" or "Follow-up: action items from Thursday's standup."
4. **Draft the email body with clear structure.** Follow a proven structure: opening hook (1-2 sentences establishing relevance), body (2-3 short paragraphs delivering the core message with supporting evidence or context), and closing with a single, unambiguous call to action. Keep paragraphs under 3 sentences. Use bullet points for lists of items or benefits. Avoid walls of text.
5. **Add signature and formatting polish.** Include an appropriate sign-off ("Best regards," for formal, "Thanks," for casual, "Cheers," for friendly) followed by the sender's name, title, and contact info. Review the draft for grammar, spelling, and readability. Ensure the email can be scanned in under 30 seconds.
6. **Review against checklist and deliver.** Verify the email meets these criteria: subject line is compelling, tone matches the audience, the CTA is clear and singular, the email is under 200 words for outreach or under 300 words for detailed responses, and no placeholder tokens remain unresolved. Output the final email in a ready-to-send format.
## Usage
Provide the agent with the email type, recipient details, purpose, key points to include, and desired tone. The agent will return a complete email with subject line, body, and signature.
**Prompt format:**
~~~
Draft a [type] email to [recipient details].
Purpose: [what you want to achieve]
Key points: [information to include]
Tone: [formal / casual / empathetic / urgent]
~~~
## Examples
### Example 1: Cold Outreach Email
**Input:**
~~~
Draft a cold outreach email to Sarah Chen, VP of Engineering at Cloudstream Inc.
Purpose: Introduce our API monitoring platform and book a 15-minute demo.
Key points: Their engineering blog mentions scaling challenges; our tool reduces incident response time by 60%.
Tone: Professional but conversational.
~~~
**Output:**
~~~
Subject: Reducing incident response time at Cloudstream
Hi Sarah,
I read your recent post on scaling Cloudstream's microservices architecture — the challenges you described around observability gaps resonated with several of our customers who faced similar growing pains.
We built Watchdog specifically for engineering teams navigating that transition. Our API monitoring platform provides real-time anomaly detection and automated root-cause analysis. Teams using Watchdog have cut their mean incident response time by 60%, which typically saves 10+ engineering hours per week.
Would you be open to a 15-minute walkthrough? I can show you how Watchdog would map onto Cloudstream's stack. Here's a link to grab a time that works: [calendar link]
Best regards,
Alex Rivera
Account Executive, Watchdog
alex@watchdog.io | (555) 234-5678
~~~
### Example 2: Customer Support Response to a Complaint
**Input:**
~~~
Draft a customer support response to Marcus Webb who is upset about being double-charged for his subscription renewal.
Purpose: Acknowledge the error, confirm the refund, and retain the customer.
Key points: Refund of $49.99 will appear within 3-5 business days; root cause was a billing system migration.
Tone: Empathetic and reassuring.
~~~
**Output:**
~~~
Subject: Your refund for the duplicate charge has been processed
Hi Marcus,
I'm truly sorry about the duplicate charge on your account — I completely understand how frustrating that must be, and I want to assure you that we've already taken care of it.
We've processed a full refund of $49.99 to your original payment method. You should see the credit reflected within 3-5 business days. The duplicate charge was caused by a one-time issue during our billing system migration, and our engineering team has deployed a fix to ensure this doesn't happen again.
Your subscription remains fully active and unaffected. As a thank-you for your patience, I've also applied a 15% discount to your next renewal.
If you have any other questions or concerns, please don't hesitate to reply to this email — I'm happy to help.
Warm regards,
Jamie Torres
Customer Support, BrightSaaS
~~~
## Best Practices
- **One call to action per email.** Multiple CTAs dilute urgency and reduce response rates. Choose the single most important next step and make it unmissable.
- **Front-load the most important information.** Recipients often scan only the first two sentences. Lead with relevanAudit web interfaces against WCAG 2.1 AA/AAA standards, identify violations, and produce actionable remediation reports with code fixes. Use when the user requests accessibility testing or provides relevant inputs for this workflow.
Design reproducible evaluations for AI agents with representative task sets, explicit rubrics, appropriate graders, baselines, regression gates, and failure analysis. Use when defining agent quality, comparing prompts or models, validating a release, measuring tool-use reliability, investigating regressions, or deciding whether an agent is ready for production.
Design privacy-aware observability for AI agents using traces, spans, structured events, metrics, cost attribution, dashboards, alerts, and investigation workflows. Use when instrumenting an agent, debugging intermittent tool or model failures, defining service-level objectives, analyzing latency or spend, auditing agent decisions, or preparing production monitoring.
Plan, execute, document, and retest authorized security assessments of AI agents and multi-agent workflows using safe adversarial cases, synthetic identities, canaries, and evidence-based findings. Use when defining red-team rules of engagement, assessing prompt injection or excessive agency, testing tool and identity boundaries, evaluating memory or cross-agent attacks, scoring a campaign, or verifying remediation in an approved environment.
Generate comprehensive marketing analytics reports by collecting KPIs, analyzing trends, and delivering actionable insights with attribution modeling and funnel analysis. Use when the user requests analytics reporting or provides relevant inputs for this workflow.
Design RESTful APIs with proper resource modeling, HTTP method semantics, status codes, pagination, versioning, and documentation. Use when the user requests api design or provides relevant inputs for this workflow.
Integrate with external APIs using REST clients, webhook consumers, SDK wrappers, and polling patterns with proper authentication, error handling, and retry logic. Use when the user requests api integration or provides relevant inputs for this workflow.
Create and manage budgets with variance analysis and departmental allocation. Use when the user requests budget planning or provides relevant inputs for this workflow.