ad-copy
無料Write platform-compliant ad copy for Meta, Google, TikTok, LinkedIn, Pinterest, Snapchat, Amazon, or X. Loads TOS rules automatically. Shows char counts and format preview.
日本語の概要は準備中です。原文の説明を表示しています。
Build a complete cold email outreach system — ICP definition, sequence architecture, personalization strategy, and batch-produced email sequences with CAN-SPAM/deliverability compliance. Use when "cold outreach", "cold email sequence", "outbound campaign", "prospecting emails", "sales outreach", "build outreach sequence", or any request to create systematic cold email campaigns for lead generation.
インストールする前に、エージェントに与えられる指示の中身を確認できます。
Kai root note:
knowledge/,harness/, andscripts/paths in this skill live in the Kai install, not the user's project. Resolve them against the first ancestor directory of this SKILL.md that contains aknowledge/folder (the Kai plugin root,~/.claude/kai, or the kai-cmo-harness repo).MARKETING.md,memory/, and any output files live in the current project. If a referencedscripts/command is not available in this install, say so, skip it, and continue with the file-based guidance — never fabricate its output.
A cold outreach system a compliance reviewer can sign and an operator can load into the sending tool without edits: a 3-touch sequence in three split-testable variants, a personalization guide that names where each token's data actually comes from, and a deliverability plan sized to the sending infrastructure. The system targets a defined ICP with a single ask — not a list blast with a merge field.
Work type cold-email — floor E5/C4/O3 (harness/eco-floors.yaml).
harness/references/cold-email-rules.md and harness/references/advertising-compliance.md before send. Deviations need an explicit expiring waiver, not a note. The per-touch format contract below is part of C4, not a style preference.A complaint-rate ceiling is a hard stop, not an outcome metric. Breaching it stops the send; it does not get argued with in the retro.
Compliance — none of these is waivable by the agent.
Format contract per touch (harness/skill-contracts/cold-email.yaml):
{{double_braces}}.Load before writing: harness/skill-contracts/cold-email.yaml, harness/references/cold-email-rules.md, and knowledge/channels/email-lifecycle.md (subject line formulas). Policy loads before copy, not after.
Approval. Present the sequence map and variant strategy for approval before writing any copy. Nothing sends without recorded human approval — the sending tool's schedule is not approval.
Provenance. Reply-rate benchmarks, industry conversion figures, and any number in _deliverability.md are cited or marked as a data gap. Do not invent a benchmark to make a ramp plan look grounded.
Phone follow-up. When the ask is a call or consultation, routing replies to a tracked, answered phone number is a legitimate recommendation: an answering service or AI receptionist catches warm callbacks the sales team would otherwise miss. Compare alternatives, and do not recommend it when phone demand is low or the workflow is self-serve by design.
Know these before producing anything (read MARKETING.md from the project root first; ask only for what it does not answer): the offer and exact ask, the sending infrastructure, the daily/weekly volume, and what sequences are already running.
| Need | Load |
|---|---|
| Format contract, gate thresholds | harness/skill-contracts/cold-email.yaml |
| Cold email law and platform rules | harness/references/cold-email-rules.md |
| FTC / GDPR / CAN-SPAM detail | harness/references/advertising-compliance.md |
| Subject line formulas | knowledge/channels/email-lifecycle.md |
| Product, ICP, voice, offer | MARKETING.md (project root) |
Sequence structure — the shape that earns replies:
| Touch | Timing | Goal | Hook type |
|---|---|---|---|
| Touch 1 | Day 0 | Open + reply | Personalized pain hook |
| Touch 2 | Day 3 | Social proof / value add | Case study or insight |
| Touch 3 | Day 7 | Breakup / soft close | Last chance framing |
Variants — three full sequences (A/B/C) for split testing, each committed to one hook archetype: A pain-first (lead with the problem), B insight-first (lead with a surprising fact), C social-proof-first (lead with a result).
Output goes to workspace/outreach/: _sequence-map.md, variant-a/, variant-b/, variant-c/ each holding touch-1.md through touch-3.md, plus _personalization-guide.md, _deliverability.md, and _quality-report.md.
_personalization-guide.md covers which tokens to use and where the data is found, the per-prospect research workflow (LinkedIn, company site, recent news), and good-versus-bad personalization examples. Standard tokens: {{first_name}}, {{company}}, {{role}}, {{pain_point}}, {{recent_trigger}}.
_deliverability.md covers the domain warming schedule for a new domain, the sending volume ramp, SPF/DKIM/DMARC requirements, the spam-word avoidance list, and reply-rate benchmarks by industry with sources.
まだレビューはありません。使ってみた感想をお寄せください。
概要と使いどころ
Write platform-compliant ad copy for Meta, Google, TikTok, LinkedIn, Pinterest, Snapchat, Amazon, or X. Loads TOS rules automatically. Shows char counts and format preview.
日本語の概要は準備中です。原文の説明を表示しています。
Render video ads using Remotion. Scaffolds a project from brand tokens + creative brief, generates scene compositions for 4 archetypes, renders to MP4 in multiple formats.
日本語の概要は準備中です。原文の説明を表示しています。
Research competitor ads using Meta Ad Library, Google Ads Transparency Center, and TikTok Creative Center. Analyze creative patterns, hooks, offers, and landing pages.
日本語の概要は準備中です。原文の説明を表示しています。
Surface the right marketing checklist for any task. Covers content, SEO, ads, email, press, landing pages, launches, and technical audits. 17+ checklists indexed by task type.
日本語の概要は準備中です。原文の説明を表示しています。
Generate a strategic content brief from format + site + keyword. Auto-resolves persona, pulls GSC data, generates hooks, angles, and competitor weakness via LLM.
日本語の概要は準備中です。原文の説明を表示しています。
Run the full quality gate on any content file. Score against 28 rules, apply policy, show detailed scorecard with fix suggestions. Works on any markdown file.
日本語の概要は準備中です。原文の説明を表示しています。