X Long-Form Post is a prompt-based writing skill. It sets voice rules (short declarative sentences, contrarian hooks backed by real numbers), a post structure from hook to payoff, and a requirement that every post carries at least one ASCII diagram in a code block.
Before the draft is returned, the agent checks it against a banned vocabulary list and 24 AI-writing patterns, then scores it from 100 with a humanizer rubric. A voice template lets you describe your own style.
What You Get
- A complete X article or thread with a hook, setup, sections, an uncomfortable truth and a payoff
- At least one mobile-width ASCII diagram: flow, before and after, hierarchy or a block-character bar chart
- A draft cleaned against a banned word list and 24 AI-writing patterns
- A humanizer score band that tells you whether to ship or rewrite
- A voice template you fill in with your own sentence style, signature moves and example post
Common Use Cases
- Turning a debugging story or business metric into a founder post
- Checking an existing draft for AI-sounding phrasing before publishing
- Adding an ASCII diagram to explain a system in a text-only post
- Logging each post and comparing it to a baseline cohort after publishing
What Makes It Unique
The skill enforces a written checklist, including a ban on negation-based contrast sentences, a banned vocabulary list and limits on em dashes and bold text. It also sets a rule that every post has a diagram, and it ties changes to hooks and structure to a baseline and candidate comparison in your analytics.
Under The Hood
This skill is instructions and two reference files. It ships no scripts and its requirements file lists no packages. The agent writes the draft itself and may read your analytics if you provide access. The humanizer scoring rubric is read from a sibling content-ops skill folder, which must be installed alongside for the full check. It does not call an API or run commands of its own.