CtrlK
BlogDocsLog inGet started
Tessl Logo

build-sequence

Use when designing a finite evidence-backed outreach sequence for an eligible contact without scheduling or sending it.

61

Quality

71%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./skills/build-sequence/SKILL.md
SKILL.md
Quality
Evals
Security

Build Sequence

Purpose

Design a finite review-only sequence with explicit timing, channels, purposes, evidence, and stop conditions. Create no automation or external action.

Required inputs

  • Independently verified ELIGIBLE status and reason
  • Supplied opt-out or stop-contact status
  • Recipient, goal, approved offer facts, and dated evidence
  • Allowed channels, maximum touches, and timing constraints when supplied
  • Voice rules and prohibited claims
  • Contactability, channel rules, rate limits, and review owner for readiness

Only eligibility and opt-out status gate offline design. Mark other missing inputs and choose conservative design assumptions without claiming permission.

Workflow

  1. Check eligibility and opt-out status. Return STOP with no sequence or copy for an opt-out, stop request, or ineligible contact. Return NEEDS VERIFICATION with no sequence or copy only when either gate is missing or unverified.
  2. Inventory facts, dates, citations, uncertainties, and channel constraints. Separate account facts from personal knowledge.
  3. Set a finite touch count and end. Without a supplied maximum, use the fewest touches needed and label the count a design assumption.
  4. Define every touch with sequence number, channel, relative delay, purpose, evidence IDs, and optional copy. Vary purpose rather than repeating the pitch. Label every copy block DRAFT FOR HUMAN REVIEW.
  5. Add pre-touch gates. Before each touch, recheck eligibility, opt-out or stop-contact status, replies, contactability, current channel policy, rate limits, and human approval.
  6. Stop downstream touches after any opt-out, negative reply, ineligibility, invalid contact, or final touch. Pause for any other reply and route it to human review; never branch automatically.
  7. Mark NOT READY TO SEND when permission, policy, rate-limit, or approval evidence is missing. Do not withhold an otherwise allowed offline design.

Output format

Return drafting-gate status and reason first. When both gates pass, return:

  • Send readiness with missing checks
  • Sequence assumptions and finite end
  • Evidence ledger
  • Touch table with number, channel, delay, purpose, evidence IDs, and pre-touch gate
  • Any copy as DRAFT FOR HUMAN REVIEW
  • Terminal stop and pause rules

For STOP or NEEDS VERIFICATION, return the reason and no sequence or copy.

Guardrails

  • Never invent facts, familiarity, urgency, outcomes, credentials, consent, contactability, citations, channel permission, or personal involvement.
  • Never target from protected traits or unnecessary sensitive data.
  • Never create endless loops, automatic channel switching, or opt-out evasion.
  • Never schedule, send, browse, use credentials, or mutate an external system.

Quality check

Confirm drafting gates passed; every touch has channel, delay, purpose, evidence, and a pre-touch gate; the sequence ends; all copy is labeled; missing readiness inputs do not block offline design; every terminal condition cancels downstream touches; and no action is implied.

Example invocation

Use build-sequence for eligible Avery Lin with supplied no-opt-out status and dated Meridian Forge evidence. Return a finite offline sequence with readiness and stop rules.

Repository
llaskin/AI-SDR-Skill-Pack
Last updated
First committed

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.