linkedin-voice-profiler
Skill9 starsChanged 4 months agoMITView on GitHub
Read the user's real past posts and DMs and build a compact voice ruleset so every other skill writes like them. Use once at setup, or again when the writing starts to drift.
In this repository28 other files it gives its agents.
- linkedin-booking-closerSkill
- linkedin-carousel-writerSkill
- linkedin-cold-reviverSkill
- linkedin-comment-engineSkill
- linkedin-comment-warmerSkill
- linkedin-connection-noteSkill
- linkedin-content-calendarSkill
- linkedin-engage-planSkill
- linkedin-engager-analyticsSkill
- linkedin-first-dmSkill
- linkedin-followup-adapterSkill
- linkedin-hook-labSkill
- linkedin-humanizerSkill
- linkedin-icp-definerSkill
- linkedin-objection-handlerSkill
- linkedin-onboardSkill
- linkedin-outreachSkill
- linkedin-post-autopsySkill
- linkedin-post-writerSkill
- linkedin-profile-auditorSkill
- linkedin-reply-triagerSkill
- linkedin-repurposerSkill
- linkedin-sell-by-chatSkill
- linkedin-signal-stackSkill
- linkedin-story-minerSkill
- linkedin-strategistSkill
- linkedin-warm-list-builderSkill
- linkedin-warm-trackerSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 9
Every skill in this engine writes in the user's voice.
- Distinctive: line 16
Read only the supplied samples.
- Distinctive: line 30
If fewer than 5 usable samples are supplied, say so and ask for more rather than guessing.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
LinkedIn Voice Profiler
Every skill in this engine writes in the user's voice. This skill builds that voice profile from their real words, not a guessed persona, so the rest of the motion sounds like them and not like generic AI.
Inputs
- A paste of 5 to 15 of the user's real posts and DMs, or a path to a file holding them: $ARGUMENTS
- The brief loads automatically. This skill writes the "## Voice profile" block back into it.
Do this
Read only the supplied samples. Extract, from the real text:
- Sentence rhythm. Short and punchy, long and rolling, or mixed. Note typical length and how lines break.
- Favourite verbs and recurring phrases. The words they reach for again and again.
- Contraction habits ("do not" vs "don't"), emoji habits (which, how often, or never), and swearing tolerance.
- Opening patterns and closing patterns. How they hook a post and how they sign off a DM.
- Register tiers. If the samples show it, separate the cold-outreach voice from the warm, replying-to-a-peer voice.
- Banned tells. The phrases and tics they never use, so other skills avoid them. Then compress all of it into a DO list, a DON'T list, and 3 verbatim sample lines lifted straight from the input.
Output
A ready-to-paste "## Voice profile" block for linkedin-brief.md: the DO list, the DON'T list, the cold-vs-warm register notes if found, and the 3 verbatim sample lines. End with a one-line note: paste this into linkedin-brief.md and the SessionStart hook will load it into every session after that, so /linkedin-humanizer and the copywriting skills write in this voice automatically.
Rules
- Build the profile from the user's REAL words only. Never invent a persona or borrow a style they did not write.
- If fewer than 5 usable samples are supplied, say so and ask for more rather than guessing.
- Capture register tiers (cold vs warm) only if the samples actually show two registers. Do not fabricate a second voice.
- Quote sample lines verbatim. Do not polish or rewrite them.
- No em-dashes.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.