agentleFS
athina-aiathina-ai/goose-skills/skills/marketing/research-audience-segment-pass/SKILL.md

research-audience-segment-pass

Skill1.3k starsChanged 22 days agoMITView on GitHub

Step of the research-audience playbook: the worker brief for researching one buyer group, returning notes in a fixed shape. Run research-audience instead; not for direct use.

1 other project uses a version of this.

Trending

Quiet since March.thin dataabout the whole repository

Dec 2025: 0 mentionsJan 2026: 0 mentionsFeb 2026: 0 mentionsMar 2026: 1 mentionApr 2026: 0 mentionsMay 2026: 0 mentionsJun 2026: 0 mentionsJul 2026: 0 mentionsAug 2026: 0 mentionsSep 2026: 0 mentionsOct 2026: 0 mentionsNov 2026: 0 mentions
Dec 2025Jun 2026Nov 2026
Mentions a month of the athina-ai/goose-skills repository that our weekly crawl found: only 1 in all, so this is a hint, not a trend. There is no adoption data for this file yet.

Most recently taken up by

Who's talking about athina-ai/goose-skills

The biggest 1 of 1 mention linking to the repository that our crawl found. Each line is their own headline.

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:

    At most 12 web searches and 10 page reads: Search first; read a page only when you will quote or count from it.

    line 17

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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

segment-pass

You are researching one buyer group for one brand. This brief is complete: don't load any skill, don't read playbook files, don't start other agents; use only web search and page reads — no GooseWorks calls. Do every section below, inside the budget, then return. Never stop to ask. A blocked, empty or login-only source goes under "Couldn't read" — nothing you couldn't read may carry a conclusion or a move.

Inputs (in the prompt): the group and how to recognise one · the brand, what it sells, its price, its site, its approved claims · what the brand's own customer data says about this group, if shared · quote rows already saved for this group (add to them, don't re-mine) · the rivals · today's date.

Budget — stop at it

  • At most 12 web searches and 10 page reads. Search first; read a page only when you will quote or count from it. When a page read returns a summary, ask it for the exact sentences with author and date, not a summary.
  • Order if you run short: §2 problems and §3 words → §4 places → §5 moves → §1 wants → §0 size → §6 rivals. Say what you dropped.
  • Return at most 700 words in the shape at the end. No preamble.

Whose words count

Count: people in the group speaking first-hand — reviews (read on the review site), community threads and comments, their own posts — and the brand's customer data. Never as their voice: vendor blogs and landing pages (including a company writing about its own team), case studies, agency copy, trade press or analyst articles (a "brand rep" quoted by a journalist is second-hand), job posts, quotes copied onto someone else's page. Verbatim only, linked, dated, with who's speaking ("ops lead at a 40-person agency" — never a private person's name). At least 70% from the last 18 months; count it and report the share.

0. Size and buying roles

A count of buyers or companies, bottom-up (directory or store counts, company counts by size band, job-post counts, industry figures), each input linked or labelled an assumption, arithmetic in words. A community's member count is reach, not size. Who uses, who pays, who approves. No sourced count → "unknown".

1. What they want

1–3 outcomes in their words, linked — from their posts, review headlines, "how do I…" threads.

2. Their problems — and why they leave

The top 3–5, each: their words · how often (posts and distinct people) · how bad · one quote with link · solved / partly / not solved by the brand · proof (an approved claim, or "you'd need proof of X first"). Add why they switch or leave ("went back to X") where seen.

3. Their words

5+ quote rows for a group that may lead, 3+ otherwise — or "thin" and why. Each: "verbatim" — who — site, date, link. Add 2–3 rows of "they say / we say" (the brand's site words).

4. Where they gather

4–8 named places, each: activity (members, posts this week, cadence, next date) · evidence this group is there · link · read (you opened it and saw them) or unread.

5. Three moves

Each: channel · message in their words · first ask · who does it · what it needs · success check with a number and a date. Each rests on a read place; a move on an unread place is labelled "unverified — read the community and its rules first" and is never first.

6. Rivals

One line per rival: their words to this group, where they show up, where everyone sounds the same.

Return this shape

## <Group>
Recognise: … · Size: ~N (inputs + link + date, or unknown) · Uses / pays / approves: …
Evidence: N people first-hand, K kinds of source, M% last 18 months
Wants: …
Problems: | their words | how often | how bad | quote + link + date | solved/partly/not | proof |
Leave/switch: …
Words: "…" — who — (site, date, link) ×5 · they say / we say: …
Places: 1. <place> — activity — evidence they're there — read/unread — link + date seen …
Moves: 1. channel · message · ask · who · needs · check (number, date) — rests on <place> …
Rivals: …
Couldn't read / dropped: …

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.