linkedin-authority-engine

Translate raw LinkedIn analytics into a clear diagnosis — what is working, what is not, and exactly 3 actions for next month. Use when the user pastes their numbers or asks "what do my analytics mean", "why did reach drop", "what should I do differently". Reads the profile goal and past performance; writes the diagnosis back to memory/.

Build a sharp, post-ready persona of the one person the client writes for on LinkedIn — title, pains, jobs-to-be-done, the words they use, what makes them save or DM. Use after the niche is defined, when content talks to "everyone", or before planning pillars/calendar. Reads the profile and writes the persona back to authority-context.md §4 after the user confirms.

Read and write protocol for the client memory substrate (authority-context.md + memory/). ALWAYS use it before generating content (read the profile + winning patterns) and at the end of every generation session (write-back). Defines the 3 gates of auto-enrichment.

Turn a topic, framework, or long-form idea into a LinkedIn carousel script — cover slide + value slides + closing CTA slide, one idea per slide, plus the post caption. Use when the user wants a carousel/document post or has a framework that earns the swipe. Humanizes the copy and hands off to the visual-brief skill for design.

Generate a 4-week LinkedIn content calendar tuned to the client's pillars, cadence, and audience — a day-by-day plan with pillar, format, topic, hook angle, and CTA goal per post. Use after pillars are defined, or when the user wants a month planned. Saves the calendar to outputs/strategy/ and hands each slot off to /guided.

Define the 3 (up to 5) recurring themes the client returns to relentlessly, each with 5-10 ready post topics. Use when content feels random, the user asks "what should I post about", or after the niche is defined. Reads the profile, grounds pillars in what already performs (memory/), and writes them back to authority-context.md §2/§11 after the user confirms.

Catalog of 22 LinkedIn content types with structure, funnel-stage mapping, and reach/conversion badges, plus a selection guide by objective and funnel stage. Use when deciding the post type before choosing the copywriting framework.

Eight copywriting structures for LinkedIn posts (AIDA, PAS, BAB, FAB, Star-Story-Solution, APP, HSO, Storytelling) with a guide for choosing by situation. Use when deciding the structural framework of a post before writing.

Diagnose why a post's closing is weak and rewrite it to maximize the one signal that matters — saves, comments, shares, DMs, profile visits, or clicks. Use when the user has a post but the ending falls flat, defaults to "Thoughts?", or doesn't match the objective. Loads the CTA bank; returns a diagnosis plus 3-5 alternatives. Inline, no file output.

A bank of LinkedIn CTAs by objective (saves, leads, engagement, followers, sales, newsletter) with formulas and mistakes to avoid. Use when choosing the closing CTA of any post to maximize the right signal in 360Brew.

Internal resource loaded by the linkedin-authority-engine:init command to run the discovery interview (Path A, in-session). Do not use on its own. Questions are grouped by the 13 sections of authority-context.md; * = question that feeds a critical profile field.

Bank of 147 LinkedIn hooks organized by type (proof of work, authority proof, transformation, contrarian, confession, curiosity, story, lists & frameworks). Use when generating post headlines, opening lines, and hook variations for any content category.

Deep discovery via Unipile API v2 — pull a client's LinkedIn profile, recent posts and company page into a dossier that pre-fills the init interview. Use when a LinkedIn URL is available and UNIPILE_API_KEY resolves, or on "pull my LinkedIn", "refresh my profile data", "audit competitor X".

Define or sharpen the client's LinkedIn niche — audience, problem, unique angle, one-line positioning. Use when positioning feels vague ("marketing leaders"), the content reads as generic, or before building content pillars. Reads the authority profile, proposes a sharper niche, and writes it back to authority-context.md after the user confirms.

Audit a LinkedIn profile (headline, banner, photo, About, Featured, Experience) and return prioritized fixes with rewritten copy, scored by impact. Use when the user pastes their profile or asks "audit my LinkedIn", "fix my headline", "is my profile converting". Grounds rewrites in the client's positioning; saves the audit to outputs/strategy/.

Convert an existing piece of content (blog post, tweet/thread, YouTube transcript, podcast, Notion doc) into a native LinkedIn post — one idea, rewritten for LinkedIn dynamics, not pasted verbatim. Use when the user pastes or links an external artifact to bring onto LinkedIn. For a raw personal experience with no external source, use story-extractor instead. Runs the full finalization pipeline and saves to outputs/posts/.

Extract a publishable LinkedIn story from a raw lived experience — structure it into a story-arc post (situation, tension, turning point, outcome, lesson) that starts at the punch. Use when the user dumps a first-person experience with no external source, says "something happened", or wants to turn a moment into a post. To repurpose an existing blog/video/tweet, use repurposer instead. Respects the profile's vulnerability level; runs the full finalization pipeline.

Ready-to-use templates for 7 LinkedIn post categories (career lesson, achievement with numbers, failure, debunk, how-to, opinion, behind the scenes) with section structure, placeholders, and filled-in examples. Use when writing posts to get a structured starting point by category.