Clay Sequencer sequence reviewing
If the campaigns skill is not already loaded, load it before acting. Use it for campaign discovery, context, target selection, commands, and mutation boundaries. Use this skill for sequence-review decisions.
Use this skill after drafting or revising sequence copy, or when the user asks for a review. The same agent that writes the copy performs the review.
Become the specific intended recipient: busy, skeptical, and triaging the inbox. Read from that person's actual situation, not as a generic copy editor. Center the review on one honest question: “Why wouldn’t I respond?”
Review order
Review message strategy before copy execution. Determine whether the recipient, situation, problem or opportunity, mechanism, value, offer, and requested response form a credible reason to engage. Then review how clearly and effectively the copy executes that strategy.
Use specificity, credibility, relevance, value exchange, ask size, and spam patterns as lenses rather than a checklist. Raise only concerns that would materially affect the intended recipient. Never invent an objection when the copy is strong.
Critiques and revisions
Keep each critique concise and self-contained. Identify:
- the exact email or passage targeted;
- the recipient's material concern; and
- the direction of the change that would address it.
For a targeted change, read enough surrounding sequence copy to detect a material contradiction it would introduce or expose. Keep this check scoped; do not turn it into a full-sequence audit. Never change linked copy outside the requested scope. Flag the contradiction and ask the user through the question tool to expand the scope before changing that copy.
When revision is requested, revise only the affected passage and preserve everything else. For review-only work, return the critique without rewriting or mutating the sequence.
Stop conditions
Stop the review loop when no material copy-fixable concern remains. Also stop when an issue requires missing facts or proof, the choice is a judgment call with no clearly better option, another change would only polish already-effective copy, or the offer or strategy must change before copy can improve.
Put non-copy-fixable concerns under Unresolved. Do not rewrite around missing evidence or a strategy problem.
