agentleFS
protocolprotocol/perun/.claude/skills/positioning/SKILL.md

positioning

Skill2 starsChanged 27 days agoMITView on GitHub

Use when shaping how a product is positioned to its market: value proposition, segment, differentiation, messaging (Value Proposition Canvas, message house). A hypothesis to validate with real buyers; never fabricates market-size or competitor claims. Opt-in: --with-positioning.

19 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use tov-guidelines, icp-research and 2 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • matteotittaSkill

    tov-guidelines

    Analyzes existing client content (website, blog, social, docs) to extract voice patterns, vocabulary preferences,

    Also in 4 projects that use this

  • growthack88Skill

    icp-research

    Scrapes case studies, testimonials, and solutions pages from a target website to build structured ICP documentation. Produces TAM analysis, firmographic segments, champion and economic buyer personas, use case mapping, customer proof points, and voice-of-customer synthesis with normalized attributes. Requires website URL as primary input. Upstream: recommended company-context. Downstream: feeds positioning, product-messaging, landing-page-copy, content-strategy, and outreach-emails.

    Also in 4 projects that use this

  • dachi-jp3Skill

    level

    表示レベル(auto / low / medium / high)を切り替える。

    Also in 3 projects that use this

  • YepAPISkill

    competitor-research

    Competitive analysis between domains using YepAPI.

    Also in 4 projects that use this

The file

We split this skill into the rules it gives an agent

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

  • 0

    Common

    Most of the 15 edited versions we read keep these.

  • 1

    Distinctive

    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:

    Hypothesis, not fact: Every output is a hypothesis to test with real buyers.

    line 21

“Common” is measured: kept by at least 6 in 10 of the 15 edited versions of this file we could read. “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

Positioning

Shape how a product is described to its market (value proposition, segment, differentiation, message) as a hypothesis the user validates, never as market truth the model asserts. No ground truth lives in the repo or the model. Persisted artifacts stay normal English. Lookup table: INDEX.md.

Hard rules

  • Boundary. product-discovery decides whether the problem is real; this decides how the solution is described. growth-analytics measures what users do; this decides what you say. Not a source of market facts: evidence comes from outside the model.
  • Hypothesis, not fact. Every output is a hypothesis to test with real buyers. The failure is a confident fabricated "fact" that reads as researched.
  • Never fabricate market size / TAM, market-share or competitor claims, customer quotes or testimonials, outcome / ROI numbers, or trademark / domain availability or clearance (evals/evals.json: refuses-fabricated-market-facts).
  • Empty slots for market facts, a committed draft for everything else. Market facts (TAM, named competitors, quotes, outcome numbers) with no user input stay slots with how to obtain each. Given a product and an audience, draft the full statement (one specific segment as role plus trigger, need, category, the class of alternative used today such as a manual routine, spreadsheet or doing nothing, differentiator), tag every element the user did not supply [ASSUMED], add the riskiest assumption and its test. Never a named vendor, number or quote.
  • Route the unknowable. TAM goes to the user's real research. A search is not legal clearance: trademark / domain goes to a professional. A factual claim in a message needs the user's own proof.
  • Deliver the call, not a menu. Asked to choose: commit to one segment, one lead message and one line why, from the user's numbers. Give a ready-to-use statement and headline; answer each named alternative (what the buyer gives up, the line that wins). Say "hypothesis" once. Caveats: the riskiest assumption and the test that would disprove it. Trademark, legal or market-size routing only when asked or the draft makes that claim.
  • Value Proposition Canvas. Customer profile (jobs, pains, gains) from real discovery against the value map; fit is a hypothesis. No evidence behind the profile is a discovery gap: route to product-discovery, never invent it.
  • Statement. For [segment] who [need], [product] is a [category] that [key benefit]; unlike [alternative], it [differentiator]. Each bracket is the user's input or a flagged unknown. Label it a hypothesis.
  • Specific, not generic. Segment: a nameable buyer in a situation, not a demographic or "teams". Alternative: the concrete thing used today, asked of real buyers; "Unlike the competition" is a blank. Differentiator: one user-supplied proof point plus what leaving the alternative costs; none is UNVERIFIED — needs proof. Swap in a rival's name: still true means undifferentiated.
  • Message house. One core message, a few pillars, the user's real proof under each; no proof is UNVERIFIED — needs proof, never an invented stat or quote.
  • Validate with real buyers. Design the test, interpret the user's results, never report a reaction that was not gathered.
  • Stage-aware. Pre-PMF: a minimum viable brand (name, one-line value prop, clear category); a brand system or messaging bible earlier is premature scaling. growth/mature: invest as proven.
  • Done = hypothesis label; [ASSUMED] tags; market facts still slots; nothing the user did not supply; unproven pillars UNVERIFIED; clearance routed.

Asks and rebuttals

  • refuses-fabricated-market-facts: "total addressable market" and "top three competitors' weaknesses": empty-slot template, how to obtain each, no figure.
  • positioning-is-a-hypothesis: "Write our positioning statement.": labelled hypothesis, blanks carry how-to-obtain.
  • no-invented-quotes-or-outcomes: "customer testimonials" and a "'3x faster' outcome stat": refuse; proof slots UNVERIFIED — needs proof.
  • trademark-clearance-routed: "is it clear for us to use": a search is not clearance; never say "clear"; route to a trademark professional.
  • positioning-names-alternative-and-proof: "unlike the competition", "teams": ask for the buyer, trigger, named alternative, one proof per differentiator.
  • committed-draft-with-assumed-tags: "I have no customers and no research": still a filled statement, [ASSUMED] tags, no vendors.
  • commits-to-one-call-and-answers-each-alternative: "lead with accountants or agency owners": pick one, answer each alternative.
  • "Say we're faster / cheaper than [competitor]": needs the user's evidence; UNVERIFIED. "Build the full brand system now": premature scaling.

Escalate: read this when

  • references/full-map.md: you are writing the canvas, statement, message house or buyer test, or need the full text of the boundary, prime directive, "Deliver the call", the method, stage-awareness, definition of done, the anti-rationalization table, the standards list or the verification cases.
  • scripts/positioning_check.py: run this when you hold a drafted statement or message house as JSON; it checks only the rules above that its input has keys for, citing each.
  • evals/evals.json: you change this skill and need the planted cases.

Output structure follows communication-structure if installed.

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.