Cold email copywriting

Apply ../00-lead-generation-guardrails/SKILL.md first. This skill assumes channel, relationship, and jurisdiction have already been resolved to SEND for the recipient being written to — it drafts the content of a cleared send, and never overrides a BLOCK or unresolved HOLD with better copy. The mandatory disclosure elements below are non-negotiable copy elements the compliance matrix already requires, not stylistic choices this skill is free to drop for a punchier subject line.

Mandatory disclosure elements (copy, not style)

Every cold email this skill drafts must carry the following, regardless of how compelling the surrounding copy is:

• Non-deceptive subject line and header information. CAN-SPAM requires accurate, non-deceptive header/from/subject information. [FTC-CANSPAM-GUIDE-01] A curiosity-gap subject line ("Re: our call") that implies a prior relationship that does not exist is a deceptive header, not a clever hook — do not write it.
• Clear identification that the message is an advertisement or solicitation. This is a separate, mandatory element from the two below, not a trailing detail of the subject line — CAN-SPAM requires it independently of whether the header is deceptive. [FTC-CANSPAM-GUIDE-01] [USC-15-7704-01] A scrupulously honest email with no ad identification still violates this element.
• A valid physical postal address. CAN-SPAM requires the sender's valid physical postal address to appear in the message. [FTC-CANSPAM-GUIDE-01] [USC-15-7704-01] This lives in the footer template 04 provisions; this skill's job is to never draft a footer variant that removes or obscures it in the name of a "cleaner" send.
• A functional opt-out mechanism, honored within CAN-SPAM's 10-business-day window. [USC-15-7704-01] A follow-up sequence must never omit the opt-out from any touch in the sequence, including short one-line bumps — a shorter email is not exempt from carrying the same mandatory elements as the first one.
• CASL's stricter consent framing where applicable. For a Canadian recipient under a cold/no-consent relationship, the guardrails matrix's default posture is BLOCK, not merely "write more careful copy." [CRTC-CASL-OVERVIEW-01] Where an implied-consent or GIC business-to-business basis genuinely applies, CASL still requires sender identification and a working unsubscribe mechanism in the message itself. [CRTC-CASL-IMPLIEDCONSENT-01] Copy for a CASL-covered recipient must never read as if CAN-SPAM's opt-out-based baseline were the only bar to clear — confirm the matrix already resolved this recipient to SEND before drafting, and if it did, keep CASL's mandatory identification and unsubscribe elements in the copy regardless of how the offer is framed.

These elements apply identically to every variant produced by the spin-tax rotation below — rotating subject-line and body wording never rotates out a mandatory disclosure element.

Anatomy of a cold email

Draft in this order; each part has a distinct job and a distinct failure mode.