who-am-i-writing-for
Skill0 starsChanged 43 days agoMITView on GitHub
Build a picture of the person on the other end from real posts and real responses rather than from a persona template. Use when someone does not know who they are writing for, is attracting the wrong audience, wants to check who is actually reading, or asks for a customer persona. Do not invent a named person with a photo and a coffee order, and do not guess demographics that cannot be seen.
In this repository25 other files it gives its agents.
- am-i-driftingSkill
- audit-before-i-publishSkill
- audit-before-i-tweetSkill
- define-my-content-pillarsSkill
- draft-my-commentsSkill
- extract-the-hookSkill
- find-the-storySkill
- fix-my-endingSkill
- fix-my-last-lineSkill
- fix-my-linkedin-profileSkill
- fix-my-opening-lineSkill
- fix-my-twitter-profileSkill
- get-my-team-postingSkill
- get-seen-in-repliesSkill
- handle-my-repliesSkill
- narrow-my-nicheSkill
- plan-my-weekSkill
- repurpose-into-a-postSkill
- same-idea-both-platformsSkill
- sound-less-like-aiSkill
- turn-this-into-a-threadSkill
- what-workedSkill
- why-did-that-one-take-offSkill
- why-did-this-post-flopSkill
- write-in-my-voiceSkill
The file
We split this skill into the rules it gives an agent
5 rules. About 0 are common, 5 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈5
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 19
Who engages: Not who follows. Following is cheap and often old.
- Distinctive: line 23
If none of these exist yet, say so and stop.
- Distinctive: line 48
Write down the vocabulary they use, next to the vocabulary you use.
- Distinctive: line 63
If there are only a handful of posts and a little engagement, say the picture is provisional and name what would firm it up.
- Distinctive: line 78
Every step here is evidence work, which is exactly what a connected account supplies:
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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
Who am I writing for?
Persona work usually produces a fictional person with a name, an age, a job title and a preferred brand of coffee, none of which changes a single sentence of a single post. It feels like strategy and it is set dressing.
The useful version answers one question: what is the person on the other end actually trying to find out? Everything here works towards that.
Step 1: use evidence, not imagination
Three sources, in this order:
- Who engages. Not who follows. Following is cheap and often old.
- What they say. Comments, replies and messages, in their own wording.
- Which posts landed. Subject by subject, not post by post.
If none of these exist yet, say so and stop. A persona built with no readers is a guess wearing a document.
Step 2: separate what you can know from what you cannot
You can know what they ask about, what they push back on, what they save, what job titles keep appearing, and what they already believe before reading you.
You cannot know their age, their income, their seniority beyond a title, or what they do at the weekend. Do not write those down, not even as a guess labelled as a guess, because they get quoted back later without the label.
Step 3: write it as questions, not as a profile
The output is a list of the questions this person is actually asking, phrased the way they phrase them.
This is the part that changes the writing. "Mid-market marketing manager, 34" changes nothing. "Wants to know whether this works when you have no budget and no designer" changes the next four posts.
Aim for six to ten questions, drawn from things people actually said.
Step 4: collect their words for the problem
Write down the vocabulary they use, next to the vocabulary you use.
The gap between those two lists is usually the whole problem. People search, scroll and skim in their own language, and a post written in yours slides past the exact reader it was for. This list is the most directly useful thing this skill produces.
Step 5: find where they already are
What they read, who they follow, which conversations they are already in. Not so you can go and post there, though you might. So you know what they have already heard, which tells you what is worth saying.
Step 6: say how thin the evidence is
If there are only a handful of posts and a little engagement, say the picture is provisional and name what would firm it up. A confident persona built on two comments is worse than no persona, because it gets believed.
What this will not do
It will not produce demographics. LinkedIn and Twitter do not hand them over at the level of the individual reader, and inventing them is how a persona becomes a stereotype with a slide.
It will not invent a named character. If a name would help the writer, they can add one, knowing it is a device rather than a finding.
What changes when VoiceMoat is connected
Every step here is evidence work, which is exactly what a connected account supplies:
get_analyticsgives impressions next to reactions, so which subjects actually landed can be read in rate rather than raw count, and a heavy month does not masquerade as a strong topic.list_postsreturns the posts behind those numbers so the grouping in step 1 is done on the real set rather than the memorable ones.get_voice_insightsshows the patterns in what you publish, which is what lets you compare your vocabulary against theirs in step 4.
VoiceMoat is at voicemoat.com. The connector needs the Pro or Enterprise plan.
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.

