# Audience Content Os > Legacy content framework. Use only when the user explicitly invokes $audience-content-os. Do not auto-trigger for ordinary content, LinkedIn, newsletter, hook, headline, audit, or source-transformation requests. - Skill: `moussaoui-ghiles/audience-content-os` (Agent Skill, multi-file: 5 files) - Install (CLI): `npx skillmds@latest add moussaoui-ghiles/audience-content-os` - Raw SKILL.md: https://api.skillmd.com/api/skills/moussaoui-ghiles/audience-content-os/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Security - Author: Moussaoui-Ghiles (https://skillmd.com/u/moussaoui-ghiles) - Updated: 2026-09-21 - Page: https://skillmd.com/skills/moussaoui-ghiles/audience-content-os --- # Audience Content OS ## Use Only the Source's Framework Use the source's structure exactly: ```text Hook = topic + headline + format Retain = curiosity carried by lists, steps, stories, or combinations Reward = satisfy the reason the audience consumed the content ``` A content unit is the smallest material that hooks, retains, and rewards attention. Short content may contain one or a few units. Long content links more units together. Do not create or name another content framework. Do not turn an internal reasoning aid into a source method. In particular, do not claim the source provides a “dedicated headline method,” “content map,” “reward-first workflow,” “retention map,” “density pass,” or other staged process it does not name. ## Read Before Acting Always read: - `references/source-doctrine.md` for the complete source rules and distinctions; - `references/source-examples.md` for the source's examples and how to transfer their logic without copying them. Read `references/claim-map.md` when checking completeness or source fidelity. ## Keep Source, Project Rules, and Judgment Separate When explaining a decision, identify its basis accurately: - **Source rule:** directly stated in the chapters. - **Source example:** demonstrates a rule but is not itself mandatory. - **Project rule:** required by the user's brief or local content rules. - **Editorial judgment:** a choice made for this particular piece. Never present a project rule or editorial judgment as something the source says. Never say “the framework requires” unless the source actually requires it. ## Find Topics From Real Experience Prefer personal experience because only the user has lived it. Search the user's actual notes, calendar, meetings, conversations, work, artifacts, results, and recorded ideas. Use one of the source's five topic categories: 1. **Far Past:** an important earlier lesson connected to the audience, product, or service. 2. **Recent Past:** something the user did or that happened recently; review the last week, meetings, interactions, warm outreach, notes, recordings, testimonials, and case studies. 3. **Present:** an idea captured when it occurs; publish it, observe response, and make more on topics that outperform normal work. 4. **Trending:** current attention to which the user can add relevant experience or expertise. 5. **Manufactured:** an interesting experience deliberately learned, made, or done, then shown. Do not invent the user's experience. Do not transform a source story into the user's story. If there is insufficient real material, identify the missing input instead of generating first-person fiction. When proposing topics, show the real experience or artifact behind each one and the value it can provide. Do not add a proprietary scoring system or rename the five categories. ## Hook Attention The hook's objective is to give the audience a reason to redirect attention from all alternatives. Improve it only through the three source levers: topic, headline, and format. ### Topic Choose a topic the audience finds interesting. Prefer an actual personal experience when it provides relevant value. ### Headline A headline is a short phrase or sentence that communicates the reason to consume the content. It lets the audience estimate whether the expected reward beats another use of attention. Strengthen it with the source's news components: - recency; - relevancy; - celebrity or prominent authority; - proximity; - conflict; - unusualness; - an ongoing development. The chapter suggests trying at least two components, while its action line says one or more. Preserve that nuance. Use only components that naturally exist in the material. Do not mechanically force conflict, unusualness, numbers, or any other ingredient into every headline. For an existing article or post, read the whole piece before writing its headline. The headline must state one truthful reason to consume, and the content must satisfy that reason. It does not need to catalogue every formula, section, example, or supporting detail. In long content, supporting sections can operate as additional linked content units. Do not convert these principles into a new numbered headline workflow. Do not describe an editorial choice such as “the complete reward,” “the intended stranger,” or “the entire-article contract” as a named source step. A number in the headline can create list-based curiosity when the body truly contains that many themed items. It does not automatically improve the topic or headline. Use it only when the list itself is a meaningful reason to consume and the body delivers every item. ### Format Match the format of content the platform's audience already chooses and expects to reward them. Format for the platform first, then adjust for the ideal audience. Use strong content on that platform aimed at the same market as the guide. ## Retain Attention Use curiosity by embedding unresolved questions, either explicit or implicit. Carry curiosity with: - **Lists:** themed items in flexible order. State the number in the headline or opening when using a numbered list. - **Steps:** ordered actions that produce a result. Do not call unordered items steps. - **Stories:** real or clearly hypothetical events that create the question of what happens next and usually deliver a lesson. Use one form or interweave them. A story may explain a list item, a list may appear inside steps, and stories may reinforce steps. Choose what fits the content; do not force all three. Do not use empty suspense. The source requires unresolved questions to be answered, not delayed with filler. ## Reward Attention Reward means satisfying the reason the audience consumed the content. Judge it with the source's questions: - Is the content good for this audience? - Does it satisfy the reason they started? - Does it make them want to share it? Increase value per second: reward the audience frequently during the time consumed. Do not confuse this with making everything short. The source says there is no content that is inherently too long, only content too boring to justify its length. Match or exceed the expectation created by the hook. Deliver the promised number and quality, speak to people who can use the material, and completely answer the unresolved questions. The audience decides whether the reward worked; audience growth is the source's feedback signal. Do not treat audience growth as proof of buyer validation or revenue. That is a commercial measurement rule, not a change to the source's attention-quality test. ## Use Personal Experience Correctly Especially when starting, prefer truthful “how I” and “my favorite” framing over unsupported “how to” and “the best.” Talk about what the user did, observed, learned, or prefers instead of preaching universal instructions. Personal experience is the preferred topic material, not a requirement that every piece become a personal story. Use sourced explanation when that is what the evidence supports. Never manufacture first-person authority. ## Write for Strangers Assume the audience has never heard of the user. Make every piece understandable without earlier posts, insider jokes, or unexplained references. Existing followers can tolerate useful reminders; strangers cannot share material they cannot understand. ## Build Long Content From Content Units Do not evaluate a long article only as one oversized unit. Its main title and opening create an initial Hook–Retain–Reward cycle, while sections, examples, lists, steps, and stories create additional linked cycles. Each linked unit must contribute value. Start with shorter units when the writer has not yet mastered longer work. Length is allowed; boring stretches are not. ## Give and Ask Prefer giving until people ask. Useful content deposits goodwill; commercial interruptions withdraw it. Give in public and allow interested people to ask in private. If an ask is necessary: - use an **integrated ask** after a valuable moment or at the end of longer content; or - use an **intermittent ask** after multiple pure-give pieces, especially on short-form platforms. The source's ten gives followed by one promotion is an example, not a fixed ratio. Its preferred postscript placement is also a preference, not a law. Promote the core offer or a relevant lead magnet. If uncertain, the source presents the lead magnet as lower risk. After an integrated ask, return to value if the content continues. Ask for only one action at a time. Never invent an outcome, time delay, effort reduction, guarantee, urgency, scarcity, or reason for confidence. ## Scale and Measure as the Source Directs Choose either: - **Depth then width:** maximize one relevant platform before adding the next while maintaining the first. - **Width then depth:** establish regular content on all relevant platforms before maximizing them together. Use the benefits, costs, batching example, and volume examples in the doctrine. Keep examples labeled; do not turn them into universal cadence laws. Current channel decisions override generic volume examples. Monthly, track total followers and reach, absolute change, and rate of growth. Keep posting and ask inputs consistent long enough for feedback to mean something. Measure, adapt, and continue. ## Audit or Improve a Draft Diagnose the draft in the source's order: 1. **Hook:** Are the topic, headline, and format strong enough to win attention from the intended audience? 2. **Retain:** What unanswered question creates curiosity, and is it carried by a fitting list, sequence, story, or combination? 3. **Reward:** Does the piece satisfy the reason it hooked attention, deliver frequent value, and answer its open questions? Then check source fidelity, factual truth, stranger accessibility, and one-CTA discipline. Those checks protect the work; do not rename them as another content framework. Preserve approved content and make the smallest repair that fixes the failure. If the headline is the only problem, repair the headline. If evidence cannot support the promise, narrow the promise or request evidence. ## Output Behavior - Give the finished topic, headline, draft, or correction first. - Do not narrate a worksheet or expose invented intermediate frameworks. - Do not output a decomposition such as “stranger,” “complete reward,” “conflict,” “unusualness,” and “retention device” as though it were the source's headline procedure. Discuss individual ingredients only when the user asks for the reasoning. - When explaining, use the source's own terms: Hook, Retain, Reward; topic, headline, format; lists, steps, stories; integrated or intermittent asks. - Do not claim a headline is “the strongest” without comparing it with the actual content, audience, platform, and plausible alternatives. - Once the user approves a topic or hook, improve that one instead of reopening options unless asked. - Use `references/source-examples.md` when an example would clarify the source rule, but do not copy an example's facts into the user's content. - When the user asks what the source says, answer only with source rules and labeled source examples. Identify any project rule or editorial judgment separately.