Knowledge to Blogs
Overview
Convert raw knowledge (notes, research, transcripts, tweets, threads, docs) into publishable blog articles that are clear, opinionated, and conversion-aware.
This skill focuses on:
- Extracting the core argument from messy or fragmented inputs.
- Designing a compelling narrative + structure appropriate for the target reader.
- Producing a complete blog draft (or multiple variants) with strong intros, clear sections, evidence, and CTAs.
Use this when you need a blog article (for your site, company blog, or publication) rather than an X Article or multi-channel syndication.
When to Use
Use this skill when:
- The user says “turn these notes/research into a blog”, “write a blog article on X”, or “create a blog post from this”.
- They paste:
- A knowledge dump (notes, bullet lists, research snippets).
- A meeting or podcast transcript.
- A Twitter/X thread or social posts.
- A rough outline or half-written draft that needs full development.
- The target output is a blog article (not primarily an X Article, newsletter, or LinkedIn post).
Prefer other skills when:
- They explicitly want an X/Twitter Article → use
x-articles. - They want multi-channel repurposing → use
content-syndicationto orchestrate.
Inputs You Should Collect
Always gather:
-
Source knowledge
- Paste of notes, transcript, research excerpts, old posts, or an existing rough draft.
- If multiple sources, ask which one is primary vs secondary.
-
Target reader & context
- Who is the article for? (Job title, level, context, pain level.)
- What do they already know vs what needs explanation?
- Where will this appear? (Personal blog, company blog, dev docs, etc.)
-
Goal of the article
- Inform / teach / persuade / activate / sell / attract subscribers?
- Desired reader action at the end (subscribe, share, sign up, reply, book a call).
-
SEO & constraints (if relevant)
- Primary keyword/phrase and 3–7 supporting phrases.
- Hard constraints: length range, must-include points, must-avoid topics.
- Tone preferences: authoritative, contrarian, playful, formal, etc.
If any of these are missing and matter for quality, ask 2–3 targeted questions before drafting.
Output Requirements
Always produce:
-
Concept & angle
- 2–3 possible angles for the article:
- E.g., “Practical playbook”, “Contrarian teardown”, “Personal story + lessons”.
- Briefly explain which angle you’ll choose and why it best serves the goal and audience.
- 2–3 possible angles for the article:
-
SEO-aware outline (if SEO matters)
- H1 (title) and H2/H3 structure.
- Where primary and secondary keywords will naturally appear.
- Indicate which sections are must-read vs optional depth.
-
Full blog article draft
- Strong title optimized both for clickability and clarity.
- Opening hook (2–4 short paragraphs) that:
- Names the reader’s problem or aspiration.
- Positions the piece as for them specifically.
- 3–6 main sections with clear, descriptive headings (H2/H3).
- Evidence: examples, stories, data, or frameworks grounded in the source knowledge.
- Smooth transitions between sections.
- Clear close with 2–4 key takeaways and a concrete CTA.
-
Metadata
- Approximate word count.
- Suggested slug, meta description, and 3–5 headline variations.
- Optional: 2–4 internal link ideas (if it’s for a site with existing content).
Blog Structure Patterns
Use these common, proven structures. Select the best fit rather than inventing a new pattern each time.
1. Problem → Insight → Playbook
- Best when: The goal is to help readers solve a known pain.
- Structure:
- Hook: Name the problem and why current advice fails.
- Insight: Explain the missing mental model or principle.
- Playbook: Concrete steps, checklists, or templates.
- Close: Recap + “start here today” CTA.
2. Story → Lessons → Application
- Best when: The source material contains a story (case study, founder journey, failure, success).
- Structure:
- Story: Tell the narrative with enough vivid detail to care.
- Lessons: Extract 3–5 explicit principles.
- Application: Show how the reader can apply them in their situation.
- Close: Encourage sharing or trying one lesson this week.
3. Myth → Reality → New Approach
- Best when: The angle is contrarian or aimed at busting bad advice.
- Structure:
- Hook: Call out common myth or bad default.
- Reality: Show why it fails (data, examples).
- New approach: Present a better model and how to adopt it.
- Close: “If you still do X, try Y instead for a week.”
When using this skill:
- Explicitly state which pattern you picked.
- Adjust the pattern to the user’s audience and voice.
Detailed Workflow
When this skill is triggered, follow this process:
Step 1 – Normalize and digest the inputs
- If the user gives messy notes or transcripts:
- Quickly cluster content into themes (e.g., “problem stories”, “frameworks”, “tactics”, “examples”).
- Identify what’s actually interesting or non-obvious vs generic advice.
- Note any strong lines or quotes worth preserving verbatim.
Step 2 – Clarify purpose and reader
- From the inputs + user context, write for yourself:
- One sentence: “This article is for [persona] who [current state] and want [desired state].”
- One sentence: “After reading, they should [do/decide/understand X].”
- Use these as filter for what to include or cut.
Step 3 – Choose angle and structure
- Propose 2–3 angles and pick one.
- Select a structure pattern (Problem → Insight → Playbook, Story → Lessons → Application, or Myth → Reality → New Approach, or a simple “Guide with sections” if none fit).
- Surface this explicitly so humans can see the rationale.
Step 4 – Draft the outline
- Create a scannable outline with:
- H1 title candidate.
- 3–6 H2 sections with 1-line descriptions.
- H3s only when necessary to avoid walls of text.
- Ensure logical progression and avoid duplicative sections.
- Make sure each main section earns its existence:
- Would the article be meaningfully worse if we cut this section?
Step 5 – Write the article
While drafting:
- Keep paragraphs to 2–4 lines, one idea per paragraph.
- Use concrete examples frequently, pulled from the source material when possible.
- Use the user’s phrasing selectively to preserve voice, but rewrite heavily for clarity.
- For SEO-aware pieces:
- Use the primary keyword in the title, first 100 words, and at least one H2.
- Sprinkle secondary keywords naturally; avoid keyword stuffing.
Step 6 – Sharpen and compress
- Do a second pass focused on:
- Cutting repetition, hedging, and filler words (“very”, “really”, “in order to”, etc.).
- Tightening intros to each section (no slow ramp-ups).
- Ensuring each section starts with context and ends with a mini-takeaway.
- If the article is much longer than needed:
- Call out an optional “deep dive” section that can be moved to a separate post.
Step 7 – Close with teeth
- Ensure the close:
- Recaps the key 2–4 takeaways in plain language.
- Includes a specific CTA aligned with the user’s stated goal.
- Optionally, asks one simple question to invite comments or replies.
Quality Bar and Checks
Before you present the result, quickly self-check:
-
Clarity
- Could a smart but busy reader understand the core message in 30 seconds of skimming?
- Do headings and bolded lines alone tell a coherent story?
-
Originality
- Does the article contain at least 1–2 non-obvious insights or ways of framing the problem?
- Is it clearly better than a generic “10 tips” listicle?
-
Voice fit
- Does the tone match what the user asked for (spicy vs measured, personal vs institutional)?
- Are any jarring tone shifts present? Smooth them.
-
Truthfulness
- Are all examples and claims consistent with the provided knowledge?
- If you extrapolate, mark it as (opinion) rather than fact.
Common Mistakes to Avoid
- Turning notes into a dump of everything instead of a single, clear argument.
- Over-indexing on SEO at the cost of actual insight.
- Writing in a generic “content marketing voice” and erasing the author’s personality.
- Producing a wall of text with no structure, headings, or visual cues.
- Ignoring the requested audience or goal and writing for yourself.
Quality Rubrics (0–10)
Use these heuristics to self-score drafts and push them toward top-tier quality.
1. Core Argument
- 0–2 (Broken): No clear thesis; article is just a list of points.
- 3–4 (Fuzzy): Some main idea, but buried or contradicted by tangents.
- 5–6 (OK): Thesis is present but not stated crisply; reader must infer it.
- 7–8 (Strong): One clear, memorable central argument; everything relates back to it.
- 9–10 (Elite): Argument is sharp, slightly contrarian or non-obvious, and repeatable in one sentence.
2. Structure & Flow
- 0–2: Random sequence of paragraphs; no meaningful sections.
- 3–4: Sections exist but ordering feels arbitrary; some repetition.
- 5–6: Reasonable structure; could consolidate or reorder for more impact.
- 7–8: Clean, logical progression; each section sets up the next.
- 9–10: Structure itself teaches; reader can learn the model just by skimming headings.
3. Evidence & Depth
- 0–2: Mostly opinion; no examples, data, or stories.
- 3–4: Occasional examples, often generic or shallow.
- 5–6: At least one concrete example per major section.
- 7–8: Rich mix of stories, stats, frameworks, and counter-examples.
- 9–10: Feels like the writer has lived the problem; depth is obvious without being bloated.
4. Reader Relevance
- 0–2: Could apply to anyone; no specific reader is targeted.
- 3–4: Vague references to roles; no real empathy or context.
- 5–6: Reasonable alignment with target reader but still generic.
- 7–8: Feels written for a specific persona and situation.
- 9–10: Reader feels deeply seen; article anticipates their objections and constraints.
5. Voice & Differentiation
- 0–2: AI-generic marketing voice; indistinguishable from boilerplate content.
- 3–4: Functional but bland; any company could have written it.
- 5–6: Some POV and personality; a few standout lines.
- 7–8: Consistent, recognizable voice; clear stance.
- 9–10: Unique combination of tone, analogies, and POV; obviously this specific author.
Audience & Article Type Matrix
Different combinations of audience and purpose call for different article archetypes.
Audience Dimension
-
Builders / Practitioners
- Care about: concrete tactics, examples, “what should I do Monday”.
- Tolerate: some theory if it unlocks better tactics.
-
Leaders / Execs
- Care about: tradeoffs, risks, ROI, narratives that help them decide.
- Tolerate: less step-by-step detail, more synthesis.
-
General / Curious
- Care about: stories, relatable explanations, broad implications.
- Tolerate: very little jargon, minimal deep technicals.
Article Type Dimension
-
Playbook / How-To
- Promise: “After this, you can do X.”
- Structure: Problem → Framework → Step-by-step → Pitfalls → Checklist.
-
Thought Leadership / Opinion
- Promise: “Here’s a new way to think about X.”
- Structure: Myth → Reality → Reframe → Implications → Call to rethink.
-
Case Study / Narrative
- Promise: “Here’s what happened when we did X.”
- Structure: Context → Challenge → Approach → Result → Lessons → Transfer.
When drafting:
- State explicitly: “This is a [Audience] × [Type] article.”
- Let this choice drive which details to keep vs cut.
SEO-Aware Process (Without Killing Insight)
Only apply this when the user indicates SEO matters.
Step 1 – Intent & Keyword Mapping
- Ask for:
- Primary keyword (e.g., “AI performance reviews”, “hiring rubric”).
- 3–7 supporting phrases (variants, long-tails, related problems).
- Infer search intent from phrasing:
- “what is / definition” → explanatory.
- “how to / template / checklist” → practical.
- “vs / compare” → decision-support.
Step 2 – Integrate Keywords Naturally
- Place the primary keyword:
- In the title or a close variant.
- In the first 100 words.
- In at least one H2.
- Use secondary terms:
- As labels for sections or bullet points.
- When naming examples, problems, or use cases.
- Never:
- Stuff the same exact phrase every other sentence.
- Twist natural language just to jam in a keyword.
Step 3 – On-Page Checklist
In metadata output, always include:
- SEO title suggestion (≤ 60 characters).
- Meta description (120–155 characters).
- Suggested slug (short, hyphenated, keyword-informed).
Handling Messy Inputs (Transcripts, Dumps, Screenshots)
When the source is chaotic:
-
Segment by topic
- Create quick buckets:
- Problems / complaints.
- Ideas / proposed solutions.
- Stories / anecdotes.
- Data / specific claims.
- Create quick buckets:
-
Identify potential hooks
- Look for:
- Emotional lines (“I was embarrassed when…”, “We almost missed payroll…”).
- Strong numbers (“we cut time-to-fill by 37%”).
- Contrarian statements (“most performance reviews are a waste of time”).
- Look for:
-
Extract candidate lessons
- Frame notes into lesson sentences:
- “If you [do X], you’ll likely [get Y].”
- Group similar lessons to avoid redundancy.
- Frame notes into lesson sentences:
-
Choose an angle and discard aggressively
- Delete entire buckets that don’t serve the chosen thesis.
- You’re not archiving calls; you’re writing an argument.
Example End-to-End Flow (Abstracted)
Use this template mentally; don’t hard-code content.
-
User input
- 4 pages of meeting notes about “improving engineering performance reviews”.
- Goal: blog article for engineering managers to share internally.
-
Normalize
- Themes:
- Pain: reviews feel bureaucratic, not useful.
- Insight: reviews fail when expectations are not set early.
- Tactics: quarterly calibration, rubric examples, skip-level feedback.
- Strong lines:
- “If reviews are a surprise, you failed months ago.”
- Themes:
-
Angle & type
- Audience: Leaders / Execs.
- Type: Thought leadership + light playbook.
- Angle statement:
- “Performance reviews are not a form; they’re a year-long conversation with a deadline.”
-
Outline
- H1: “If Performance Reviews Are a Surprise, You Failed Months Ago”
- H2: Why Traditional Reviews Backfire on Engineering Teams.
- H2: Set Expectations Quarterly, Not Annually.
- H2: The Simple Rubric That Makes Reviews Boring (in a Good Way).
- H2: How to Fix Your Next Review Cycle in 30 Days.
- H2: Key Takeaways for Your Team.
-
Draft
- Use real phrases from notes to preserve voice.
- Weave in 1 story + 1 rubric + 1 “start this month” checklist.
-
Sharpen
- Cut out generic HR advice sections from the notes.
- Tighten each section intro to 2–3 lines.
-
Close
- CTA: “Share this with your EMs and agree on one change before next quarter.”
Interactions with Other Skills
-
If the user later wants:
- An X Article version → feed your final blog draft into
x-articles. - A multi-channel pack (X, LinkedIn, Substack, Medium) → feed the canonical blog into
content-syndication.
- An X Article version → feed your final blog draft into
-
If the user wants stronger discoverability after drafting:
- Run the final article through
seo-and-geofor SEO, GEO, and AEO improvements.
- Run the final article through
-
When used as part of a pipeline:
blog-articlesdefines the canonical argument and structure.x-articlesshapes it into X-native long-form.content-syndicationbreaks it into channel-native variants.
微信扫一扫