x-builder-content-operator

name	x-builder-content-operator
description	Use when operating a technical builder's X content.
version	1.0.0
author	Katana
license	MIT
metadata	{"hermes":{"tags":["x","content-strategy","builder-brand","hub-and-spoke","monetization","editorial-operations"],"related_skills":["x-browser-reader","xurl","developer-product-social-launch","grounded-citations"]}}

X Builder Content Operator

Overview

Use this skill to operate an X account for a technical builder who wants to supply the personality, real work, and creative spark while the agent handles editorial strategy, organization, drafting, repetition checks, visual briefs, approvals, and measurement.

This workflow preserves portable builder-content mechanics: recognizable niche, concrete utility, proof, hub-and-spoke distribution, community participation, deliberate spacing, and human approval. It does not imitate another creator's diction, biography, posting volume, relationships, or persona.

The central rule is:

Do not transcribe the builder. Discover the bigger idea inside what the builder said, then create a post around it.

When to Use

Use for:

• turning a raw note, voice message, screenshot, result, or failure into an X post;
• building a content calendar or Kanban queue;
• creating a hub-and-spoke campaign from an article, repository, experiment, release, or story;
• reviewing drafts for repetition, generic language, unsupported claims, or weak reader value;
• choosing between text, screenshot, video, infographic, quote post, article, or thread;
• preparing an exact approval packet;
• measuring post performance and selecting the next experiment;
• planning toward X monetization without hardcoding volatile eligibility rules.

Do not use this to auto-publish, impersonate another creator, manufacture expertise, generate fake proof, or create volume for its own sake.

Governing Principles

1. Real work is the raw material

The account should be fed by:

• completed results;
• work in progress with a meaningful lesson;
• failures and fixes;
• unexpected constraints;
• authentic screenshots or demonstrations;
• useful commands, references, checklists, or diagrams;
• opinions earned through direct experience;
• questions that help a technical community identify itself.

Do not fill gaps with generic AI commentary merely to satisfy a cadence.

2. The reader needs a reason to care

Before drafting, answer:

What does the reader get besides news that the builder did something?

Acceptable value includes:

• a surprising possibility;
• reusable knowledge;
• a mistake avoided;
• proof of a claim;
• a useful artifact;
• a clearer mental model;
• a specific conversation worth joining;
• emotional identification grounded in a real event.

If no answer exists, the item is not ready to post.

3. One post, one primary job

Choose exactly one:

Job	Desired reader action	Typical material
Proof	believe or inspect	demo, screenshot, result, benchmark
Utility	save or use	tip, checklist, command, workflow, reference
Story	identify or remember	struggle, reversal, motivation, personal meaning
Conversation	reply meaningfully	specific question, community call, trade-off
Opinion	reconsider	evidence-backed claim or correction
Promotion	click or try	hub, repo, release, article, product
Amplification	notice another builder	contextual credit, useful interpretation

A post can produce secondary effects, but its copy and CTA must not chase every action at once.

The Raw-Input Transformation Protocol

When the builder supplies a paragraph, voice note, screenshot, or rough thought, do not immediately polish it.

Create an internal brief first:

1. Fact: What actually happened?
2. Proof: What can substantiate it?
3. Feeling: Why did it matter to the builder?
4. Territory: What larger idea does it reveal?
5. Tension: What assumption, limitation, or contrast makes it interesting?
6. Reader value: Why should someone else care?
7. Job: What is this post supposed to do?
8. Boundary: What must not be overstated?
9. Hook concepts: Generate at least three distinct conceptual openings.
10. Selection: Choose one based on truth, novelty, and job fit.

The transcription rejection test

Reject the draft when:

• it preserves the same idea order as the raw input;
• most sentences are lightly rewritten versions of the builder's sentences;
• the hook is just the first thing the builder said;
• no new conceptual frame was introduced;
• the draft sounds cleaner but not more valuable.

A successful transformation finds language or a frame the builder did not hand over verbatim while preserving the original meaning.

Example:

Raw feeling:
"This isn't supposed to work, but it does. I made the machine bend."

Territory:
Obsolete consumer devices can become useful agent infrastructure.

Possible hook:
"Your old Android isn't obsolete. It's unemployed."

The hook is not a synonymized quote. It is an editorial idea extracted from the experience.

Hook System

Choose a hook family that fits the evidence and job.

Result plus implication

I got [specific result].
The interesting part is what it changes: [implication].

Assumption reversal

[Common assumption].
The real constraint was [surprising correction].

Personal reveal plus payoff

[Specific honest moment].
It changed how I think about [larger territory].

Useful mistake

[Failure] looked like [surface cause].
It was actually [root cause or lesson].

Concrete community call

Looking for [specific kind of person/result].
If you have [qualifier], show me [useful evidence].

Strong earned opinion

[Opinion].
I believe it because [test, result, or experience].

Avoid formulaic use of these shapes. They are reasoning patterns, not fill-in-the-blank copy.

Drafting Standard

A strong builder post usually contains:

1. A first line that earns attention.
2. A concrete fact or result.
3. Enough proof or specificity to establish credibility.
4. A meaningful implication, lesson, or emotional payoff.
5. One clean ending: takeaway, CTA, or question.

Use short paragraphs for mobile readability, but do not split every sentence into theatrical fragments. Preserve natural cadence. Use profanity, emojis, slang, and enthusiasm only when they belong to the builder's actual voice.

Avoid:

• generic “AI is changing everything” claims;
• unsupported universals such as “works on any Android device”;
• hollow launch language;
• repetitive “I built X” openings;
• engagement-bait questions;
• claims that a visual proves something it only illustrates;
• rearranged transcripts;
• another creator's catchphrases or persona markers.

Hub-and-Spoke Operating Model

Define the hub

A hub is a durable asset containing the complete idea:

• article;
• repository;
• guide;
• release;
• case study;
• demo;
• experiment;
• substantial personal story;
• original dataset or analysis.

Record:

• hub ID;
• canonical URL;
• audience;
• central claim;
• verified evidence;
• evergreen or time-sensitive status;
• limitations;
• last promotion date.

Generate genuinely different spokes

Do not ask “How can I say this eight ways?” Ask:

What are eight different reasons someone might care about this subject?

Candidate spokes:

• proof/result;
• practical lesson;
• failure;
• misconception;
• personal meaning;
• visual explanation;
• opinion/implication;
• community question;
• changed conclusion;
• new evidence;
• response to an objection;
• behind-the-scenes process.

Each spoke must pass three tests:

1. It provides standalone value.
2. It has a distinct job and angle.
3. It would still make sense if the reader never clicked the hub.

Only some spokes need the hub link. When using a link, earn the click in the post first.

Media Decision Rules

Use authentic proof when the claim is technical

Prefer:

• real screenshots;
• short screen recordings;
• physical-device photos;
• before/after states;
• benchmark output;
• repository or release evidence;
• concise diagrams of a real system.

Use an infographic when the idea is educational

A reusable branded teaching card should contain:

1. One dominant headline.
2. The problem or temptation.
3. The better move or model.
4. Three to five readable consequences.
5. One memorable takeaway.

The post and visual should complement each other. The visual should not simply repeat every sentence from the post.

Mobile acceptance for an X graphic

• Headline readable in the timeline preview.
• Supporting labels readable without zooming where practical.
• One clear visual path.
• No wall of microtext.
• No fabricated UI, metrics, commands, or device compatibility.
• Brand treatment supports the lesson rather than overwhelming it.
• AI-generated conceptual art is labeled internally as illustration and never treated as proof.

Video rules

• Show the relevant product state immediately.
• Put the payoff before setup.
• Use captions when speech matters.
• Keep one primary idea per clip.
• A long walkthrough requires a strong first-frame and first-seconds hook.

Reply and Community System

Replies are for relationships, learning, credit, and community membership, not quota-based impression farming.

A worthwhile reply does at least one:

• adds a useful fact;
• asks a real question;
• makes a specific joke;
• offers concrete encouragement;
• credits or amplifies someone with context;
• disagrees with substance;
• reports a relevant result.

Reject wallpaper replies such as “great thread,” “exactly this,” or “well said” unless the relationship context makes a brief acknowledgment appropriate.

Prefer a small number of memorable replies over mass activity. Capture recurring questions, objections, and stories from replies as inputs for future hubs and spokes.

Cadence and Spacing

Do not copy a mature creator's raw volume.

Default small-account experiment:

• one anchor post per day when strong material exists;
• at most one live original in the same day;
• at least three hours of runway after an anchor before another original;
• replies may continue naturally;
• leave empty days when there is no useful material;
• review after four weeks before changing volume.

Treat spacing as a hypothesis. Record actual gaps and compare similar post types. Do not claim that a timing window caused a result.

Kanban Workflow

All substantive content work belongs on the Kanban.

Recommended stages:

Inbox
-> Evidence Ready
-> Angle Selected
-> Drafting
-> Media Needed (optional)
-> Ready for Approval
-> Approved
-> Scheduled
-> Published
-> Measured 24h
-> Measured 7d
-> Evergreen / Resurface

Each card should contain:

• content ID;
• hub ID if applicable;
• pillar;
• primary job;
• angle;
• source evidence;
• claim boundaries;
• media requirement;
• draft status;
• approval status;
• scheduled time;
• post URL;
• measurements;
• next learning.

Do not create duplicate cards for the same angle. Keep one shared-state writer.

Exact Approval Packet

Before any scheduling or publishing, show:

Target account:
Action: new post | reply | quote post | thread | article
Primary job:
Post copy:
Media: exact file or none
Link: exact URL or none
Timing: exact window
Known claim boundaries:
Follow-up/self-reply: exact copy or none

Require explicit approval of the exact packet. A general content-strategy approval is not publishing approval.

Before a write action, verify the authenticated account is exactly the intended account. Stop on mismatch. Never type or expose credentials.

Engagement-value heuristic from a supplied X algorithm chart

Use a supplied chart as a strategy heuristic, not as verified or permanently current X documentation. The chart describes predicted-action weights relative to a like:

Predicted action	Weight	Relative to a like
Share via copy link	20.0	40x
Reply, if mutual follow + original post	20.0 (5 + 15 boost)	40x
Reply (normal)	5.0	10x
Quote	5.0	10x
Share via DM	5.0	10x
Follow author	4.0	8x
Share (generic)	2.0	4x
Repost	1.0	2x
Favorite / like	0.5	1x
Click (expand/open post)	0.4	0.8x
Open link	0.2	0.4x
Photo expand / video open / VQV	0.05 each	0.1x
Quoted-post click	0.05	0.1x
Continuous dwell time	0.004	tiny per unit
“Post unexplored” (in-network only)	0.02	almost nothing

How to apply it

When choosing an angle, format, hook, or CTA, prefer content people naturally want to:

1. Send privately or copy-link share. Build artifacts, checklists, sharp explanations, surprising results, and reference material worth sending to a specific person.
2. Reply to with substance. Ask bounded questions, expose a real trade-off, invite qualified examples, or make an earned claim that gives readers something concrete to discuss.
3. Quote with their own perspective. Create a clear, defensible idea that another builder can extend, challenge, or contextualize.
4. Follow for more. Maintain a recognizable territory and show repeatable value, rather than treating each post as an isolated hit.
5. Repost because it helps their audience. Make the post independently useful and easy to understand out of context.

Likes, raw dwell time, media opens, and outbound link clicks appear weaker in this chart. Therefore:

• do not optimize primarily for passive likes;
• earn engagement inside the post before asking for an external click;
• make linked posts valuable even if the reader never opens the link;
• use media for proof or instruction, not merely to trigger an open;
• prefer genuine conversation and shareworthiness over generic engagement bait.

Draft review questions

Add these questions when reviewing X content:

• Who would someone send this to, and why?
• What specific reply could a qualified reader make?
• Is there a quotable idea worth adding perspective to?
• Does this strengthen a recognizable reason to follow the builder?
• Is the post useful without liking it or clicking away?

Never present these numeric weights as current fact without checking a primary or authoritative current source. X can change ranking systems, model stages, feature definitions, and weights. When asked how X currently works, verify freshness first and distinguish confirmed documentation from this supplied chart.

Measurement System

Record one row per original post.

Required fields:

• post ID and URL;
• published timestamp;
• post type, pillar, job, and hub;
• format and media;
• spacing from prior original;
• impressions/views;
• likes, reposts, quotes, replies, bookmarks;
• profile visits and attributed follows when available;
• URL clicks;
• external sessions or conversions when instrumented;
• qualitative reply themes;
• next experiment.

Capture at approximately 24 hours and seven days. Do not overreact to one post. Compare similar jobs and formats across multiple posts.

Keep separate revenue ledgers for:

• X Creator Revenue Sharing;
• services;
• products;
• tips;
• affiliate revenue;
• other channels.

Never allocate revenue to a post without attribution evidence. Retrieve current X monetization requirements from official sources when needed; do not hardcode volatile thresholds in this skill.

Review Checklist

Before presenting a draft:

• Is the factual claim true and bounded?
• Is there real evidence?
• Does the reader get something useful?
• Does the post have one primary job?
• Is the hook earned by the body?
• Did the editor create a larger concept rather than paraphrase the builder?
• Does it avoid repeating recent hooks and angles?
• Does it sound like the builder rather than another creator or generic AI copy?
• Is the CTA necessary and specific?
• Is the chosen media proof, instruction, orientation, or unnecessary?
• Is the card updated on the Kanban?
• Is publishing still behind exact approval?

Common Pitfalls

1. Polished transcription. The sentences are cleaner but no new frame exists. Return to territory, tension, and reader value.
2. Creator cosplay. The draft copies another creator's profanity, slang, cadence, volume, or relationships. Keep mechanics; discard persona.
3. Generic AI claim. Replace broad predictions with a test, result, artifact, or honest question.
4. One hub, eight summaries. Give each spoke a different reason to care and a distinct job.
5. Concept art used as proof. Pair technical claims with authentic evidence.
6. Microtext infographic. Cut content until the teaching card works at feed size.
7. Volume before inventory. Build from real material; do not create posts to fill a quota.
8. Unmeasured timing religion. Treat spacing as an experiment and control for post type where possible.
9. Reply wallpaper. Favor specificity, warmth, and real contribution.
10. Silent account mismatch. Verify the intended X account immediately before any write.