evanbaker315/attentionos/voice-profile
Claude Code and agent skill by evanbaker315/attentionos
- community
Checking that the SKILL.md file is still in the repository.
Build or refresh brand/voice.md from the owner's own on-camera transcripts, as counted evidence rather than adjectives, including the banned openers and phrases the validator reads. Use during onboarding, when new winning posts land, or when the owner says a script doesn't sound like them.
Install
npx vskill@latest install evanbaker315/attentionos/voice-profileSKILL.md
Voice profile
brand/voice.md is evidence of how the owner really talks, built only from their own transcripts. Every line in it traces to a transcript. It is the file every script is written through, and the validator reads its banned lists.
Build it
- Collect 10 to 20 of the owner's own videos with transcripts (onboarding import or the own-account sync). Never use another creator's transcripts.
- Rank them by performance against the owner's own median. Read the top third and the bottom third.
- Count, don't guess: median sentence length, how often each recurring phrase appears, what line two usually does, how they say numbers, how they end.
- Note what flopped and why. A pattern that fails for this owner is as useful as one that works.
- Fill in the sections below, then ask the owner to approve the file. Write the date and the number of transcripts at the top.
Sections
Banned lists (the validator reads these)
- Put openers under a heading containing "Banned openers" (or "Never open with").
- Put phrases under a heading containing "Never say" (or "Banned words", "Avoid").
- One entry per bullet. Quote the exact words:
- "phrase here". Unquoted bullets are cut at the first colon or parenthesis; slashes and commas split alternatives.
Brand-wide bans that also apply to captions and posts can go in brand/brand.yaml under bannedPhrases.
Refresh
When new winners land, re-rank, add their openers to the own bank (the harvest job does this automatically), and retire patterns that stopped working. When the owner corrects a script, add the correction to brand/lessons.md the same day.
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingplugin/skills/voice-profile/SKILL.md
Popularity
- Repository stars
- 99
- Repository forks
- 11
- 7-day trend score
- 0
Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/evanbaker315/attentionos/voice-profile)