CtrlK
BlogDocsLog inGet started
Tessl Logo

profile-voice

Use when deriving a reusable messaging voice profile from user-approved writing samples or explicit approved voice guidance.

64

Quality

77%

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/profile-voice/SKILL.md
SKILL.md
Quality
Evals
Security

Profile Voice

Purpose

Describe repeatable voice traits and sentence patterns without claiming to impersonate a person. Separate observed patterns, interpretations, and missing evidence.

Required inputs

  • User-approved source samples with author or brand and context
  • Explicit approved voice guidance, if supplied
  • Intended channel and audience
  • Claims and compliance constraints

Require approved samples before finalizing sample-derived patterns. Explicit guidance may support a provisional profile; label missing sample evidence and request it. Never use recipient replies as brand-voice samples.

Workflow

  1. Confirm which samples or guidance are approved and relevant to the requested channel.
  2. Extract observable traits: sentence length, syntax, vocabulary, rhythm, openings, transitions, questions, and calls to action.
  3. Label each item Observed, Inference, or Missing. Prefer patterns repeated across approved samples.
  4. Define sentence patterns as structural recipes, not copied phrases.
  5. List banned habits from explicit guidance and evidence-backed negative patterns.
  6. Transform one supplied or clearly illustrative sentence into the profile. Preserve its facts and label illustrative text as draft.

Output format

# Voice profile
## Source basis
- [approved sample or guidance; gaps]
## Voice traits
- [Observed | Inference | Missing]: [trait and evidence]
## Sentence patterns
- [repeatable structure]
## Banned habits
- [habit]
## Sample transformation
- Before: [source or illustrative draft]
- After: [rewritten draft]
- Changes: [evidence-based explanation]
## Review checks
- [check]

Guardrails

  • Do not claim to clone, imitate, or write exactly like a real person.
  • Do not infer personality, identity, emotion, or intent from sparse samples.
  • Never add facts, familiarity, urgency, results, or claims during transformation.
  • Keep facts, inferences, and missing information distinct.
  • Do not send messages, contact people, or mutate external systems.

Quality check

Confirm sources are approved, traits are traceable, sentence patterns are actionable, banned habits are explicit, the transformation preserves meaning, unsupported choices are labeled, and the profile makes no impersonation claim.

Example invocation

Use profile-voice on Northstar Relay’s approved restrained voice notes. Produce traits, short sentence patterns, banned hype and false familiarity, review checks, and one draft before/after transformation.

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.