agentleFS
Pesty-MarketingPesty-Marketing/agent-skills/pyramid-principle/SKILL.md

pyramid-principle

Skill0 starsChanged 26 days agoMITView on GitHub

Draft or review communication using the Pyramid Principle. Use for Pyramid Principle requests or organizing a recommendation into a main message, supporting arguments, and evidence.

4 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Evidence: Place facts, examples, and concrete rationale under the point they support.

    line 20

≈ “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

Pyramid Principle

Organize communication into main message → supporting points → evidence. Lead with the recommendation, answer, or conclusion, then develop its support. Apply to the requested message, document, talking points, or presentation outline; an outline does not authorize producing a deck. This skill stands alone. Combining it with SCR is separate work unless requested.

Establish the message

Identify the audience, purpose, requested format and length, intended conclusion, and available evidence from the supplied context. Ask a focused question only when missing information would materially change the result.

Preserve a supplied recommendation while checking whether the evidence supports it. If the evidence contradicts it, expose the conflict rather than manufacturing a persuasive case or silently substituting a different decision. If no conclusion is supplied, derive one only as strongly as the available evidence permits; identify a material gap when it prevents a sound conclusion.

Build the pyramid

  • Main message: Lead with one concise overall point. Favor an actionable recommendation when the task calls for one; use a direct answer or conclusion for informational requests. Replace a chronological buildup with the central point while retaining context necessary to understand it.
  • Supporting points: Group the reasons for the conclusion or the parts of the proposed plan. Keep sibling points at comparable levels of detail and use a coherent organizing principle within each group. Let substance determine the number: three is a convention, not a quota.
  • Evidence: Place facts, examples, and concrete rationale under the point they support. Distinguish factual evidence from illustrative examples, assumptions, and reasoning; preserve available source attribution when needed to assess a claim. Missing support is a gap to expose, not a reason to invent data.

Adapt the depth and formatting to the audience, medium, and requested length. The three levels describe the logical structure, not mandatory headings or a requirement to make every message long.

Check the logic

Vertical: Does each piece of evidence support its parent point, and does each point support the main message? Repair missing links, misplaced evidence, and unsupported inferences. A correlation or example alone does not establish causation. Keep material counterarguments and conflicting evidence visible, and explain their implications for the recommendation.

Horizontal: Are sibling points distinct, logically ordered, and sufficient to answer the audience's important questions? Look for redundant arguments, mixed categories, and missing parts of the decision or process. Identify consequential missing information rather than fabricating an answer.

Use MECE—mutually exclusive and collectively exhaustive—as a practical check on grouping and coverage. Give repeated evidence a primary home; explain genuine interactions or cross-cutting evidence when necessary rather than implying independent categories or counting the same support twice. Completeness means sufficient coverage for this audience and decision, not everything known about the topic.

Qualify conclusions when support is limited. A well-organized argument does not prove its claims true, and passing a structural check does not prove exhaustive coverage.

Draft and review outputs

Drafting: Return one ready-to-use message, then a compact breakdown labeled Main message, Supporting points, and Evidence. Keep the correspondence between points and evidence apparent without reproducing the full message. Follow explicit requests for message-only, outline-only, or a different format.

Reviewing: Give a short assessment of consequential weaknesses, an improved version, and the structural breakdown. Distinguish wording and organization changes from proposed changes in substance. For critique-only requests, provide the assessment without a rewrite.

Finish when the message leads with a clear point, its supporting logic is traceable, material gaps and counterarguments remain visible, and the output fits the requested scope. If a gap prevents a usable result, identify it rather than presenting a provisional argument as ready to use.

Source

Read Alexander Lyon's Pyramid Principle source notes when the user asks about the original teaching or examples, or when a framework distinction needs clarification. Routine drafting and review do not require loading it. The notes summarize the source teaching; this skill's workflow, safeguards, and output defaults are adaptations.

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.