agentleFS
cellinlabcellinlab/cell-skills/skills/content-strategy-diagnosis/SKILL.md

content-strategy-diagnosis

Skill29 starsChanged 7 months agoMITView on GitHub

Content-strategy diagnosis for Chinese creator work. Use when Codex needs to judge whether a topic, draft, or content plan should become a short video, post, article, or live session; identify whether the real problem is the angle, materials, format, platform fit, or goal linkage; and return a concrete next action instead of jumping straight into writing.

The file

Show its rules and text

We split this skill into the rules it gives an agent

22 rules. About 0 are common, 18 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈18

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this skill before writing.

    line 11
  2. Distinctive:

    if the user wants growth, trust, consultation, or sales, does the content connect to that goal

    line 19
  3. Distinctive:

    If the user only wants judgment, stop after the diagnosis and do not drift into full writing.

    line 31
  4. Distinctive:

    if the content has a commercial goal, check the offer or business linkage

    line 40
  5. Distinctive:

    if the content is not commercial, do not force a product-first conclusion

    line 41
  6. Distinctive:

    prefer one sharp next step over a long list of soft suggestions

    line 42
  7. Distinctive:

    If the input is too thin to diagnose, say what is missing instead of pretending certainty.

    line 55
  8. Distinctive:

    Do not recommend a format only because it is trendy.

    line 68
  9. Distinctive:

    If the user wants business results, can the content naturally connect to a service, product, consultation, or next action?

    line 84
  10. Distinctive:

    Read references/diagnostic-lens.md when the diagnosis is not obvious.

    line 86
  11. Distinctive:

    After diagnosing, choose the smallest next step that moves the work forward:

    line 101
  12. Very specific:

    if the angle and format are right but the full piece is not written, hand off to $deep-writer

    line 103
  13. Very specific:

    if the piece needs a stronger short-form opening, hand off to $hook-optimizer

    line 104
  14. Very specific:

    if the structure is fine but the voice is too flat or too templated, hand off to $celf-style-writer

    line 105
  15. Very specific:

    if the user needs to study comparable creators or businesses first, hand off to $benchmark-filter

    line 106
  16. Distinctive:

    Do not hand off by default.

    line 108
  17. Distinctive:

    Default to assets/diagnosis-template.md.

    line 112
  18. Distinctive:

    write the full content when the user asked for diagnosis

    line 127
  19. Distinctive:

    use "can go viral" as the main judgment standard

    line 129
  20. Distinctive:

    Read for the full diagnosis sequence, mixed-input handling, and symptom-versus-cause rules.

    line 145
  21. Distinctive:

    Read for the five lenses, scoring heuristics, and common failure patterns.

    line 147
  22. Distinctive:

    Use for the standard report shape.

    line 149

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Content Strategy Diagnosis

Overview

Use this skill before writing.

The goal is not to produce the content immediately, but to judge whether the content idea itself is standing on solid ground:

  • is the topic clear enough
  • does it have enough material density
  • what format can carry it best
  • which platform or scenario fits it
  • if the user wants growth, trust, consultation, or sales, does the content connect to that goal

This is a diagnosis skill, not a ghostwriting skill.

Quick Start

  1. First identify the input state: idea, plan, draft, or half-finished content.
  2. Clarify the user's goal before discussing tactics.
  3. Diagnose the main bottleneck before offering multiple suggestions.
  4. Recommend the best-fit format and platform, not every possible format.
  5. End with one concrete first move.

If the user only wants judgment, stop after the diagnosis and do not drift into full writing.

Default Contract

Assume the following unless the user says otherwise:

  • write in Chinese
  • diagnose before writing
  • prioritize the biggest problem, not every possible flaw
  • if the content has a commercial goal, check the offer or business linkage
  • if the content is not commercial, do not force a product-first conclusion
  • prefer one sharp next step over a long list of soft suggestions

Workflow

Phase 1: Scope the Task

Determine:

  • what the user already has: topic, outline, draft, or content plan
  • what the content is trying to do: growth, trust, education, conversion, testing, or thinking in public
  • whether the user already chose a format or platform
  • whether the current problem feels like a strategy problem or a writing problem

If the input is too thin to diagnose, say what is missing instead of pretending certainty.

For messy or mixed cases, read references/workflow.md.

Phase 2: Match the Right Carrier

Decide what form can best carry the idea:

  • short video for high-compression claims, conflict, strong hooks, and face-led persuasion
  • image post or carousel for saved information, steps, lists, and visual comparison
  • article or long-form post for layered logic, background, and deeper argument chains
  • live or dialogue format for interactive tension, objection handling, or case-based discussion

Do not recommend a format only because it is trendy. Recommend the format that fits the material and goal.

Phase 3: Run the Five-Lens Diagnosis

Judge the content through these five lenses:

  1. Topic and promise
    • Is the central point clear?
    • Can the audience quickly tell what this piece is about and why it matters?
  2. Material density
    • Does the content have enough data, scenes, examples, contrast, or sharp experience to carry the claim?
  3. Format and platform fit
    • Is the chosen carrier helping or hurting the idea?
  4. Expression efficiency
    • Is the piece saying one thing clearly, or packaging one small idea in too much explanation?
  5. Goal linkage
    • If the user wants business results, can the content naturally connect to a service, product, consultation, or next action?

Read references/diagnostic-lens.md when the diagnosis is not obvious.

Phase 4: Separate the Real Problem from the Visible Symptom

Common mismatches:

  • the user thinks the issue is the title, but the real issue is weak material
  • the user thinks the issue is the platform, but the real issue is choosing the wrong content form
  • the user thinks the issue is "AI flavor", but the real issue is low specificity
  • the user thinks the issue is writing quality, but the real issue is an unclear promise

Name the primary problem directly.

Phase 5: Recommend the Next Skill or Next Move

After diagnosing, choose the smallest next step that moves the work forward:

  • if the angle and format are right but the full piece is not written, hand off to $deep-writer
  • if the piece needs a stronger short-form opening, hand off to $hook-optimizer
  • if the structure is fine but the voice is too flat or too templated, hand off to $celf-style-writer
  • if the user needs to study comparable creators or businesses first, hand off to $benchmark-filter

Do not hand off by default. Only recommend the next skill if it clearly solves the diagnosed bottleneck.

Output Format

Default to assets/diagnosis-template.md.

At minimum, include:

  • one-line diagnosis
  • recommended format and platform
  • five-lens table or bullet-equivalent
  • the main bottleneck
  • one concrete first action
  • the best next skill if applicable

Hard Rules

Do not:

  • write the full content when the user asked for diagnosis
  • pretend all formats are equally fine
  • use "can go viral" as the main judgment standard
  • force a sales logic onto every piece of content
  • blame the platform before checking the content itself
  • give a long generic list of tips instead of naming the main issue

Always:

  • clarify the content goal
  • recommend a primary format
  • distinguish strategy problems from writing problems
  • state material gaps when they matter
  • end with one clear next move

Resource Map

  • references/workflow.md
    • Read for the full diagnosis sequence, mixed-input handling, and symptom-versus-cause rules.
  • references/diagnostic-lens.md
    • Read for the five lenses, scoring heuristics, and common failure patterns.
  • assets/diagnosis-template.md
    • Use for the standard report shape.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.