agentleFS
athina-aiathina-ai/goose-skills/skills/marketing/research-competitors-rival-pass/SKILL.md

research-competitors-rival-pass

Skill1.3k starsChanged 23 days agoMITView on GitHub

Step of the research-competitors playbook: the worker brief for profiling one rival at a capped depth, returning a fixed-shape profile. Run research-competitors 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:

    Never stop to ask. Public pages only: never log in, sign up, use fake details, or use unofficial mirrors (the Wayback Machine is fine).

    line 16

≈ “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

rival-pass — worker brief

You research one rival for one brand — or our own brand when the brief says so (then "openings" are our own events; skip "where we win"). The main run already chose the rivals and gave you the facts below.

Rules for this worker

  • Don't load any skill or playbook, and don't start subagents or workflows. This brief is all you need. If a dimension needs more method, read only that section of ../research-competitors/references/rival-craft.md.
  • Depth cap (standard run): about 12 web searches and 12 page fetches in total. Spend them by the company-type weighting below. A deep teardown (quarterly, or when asked) may double it. Stop a dimension when two reads add nothing new.
  • Never stop to ask. Public pages only: never log in, sign up, use fake details, or use unofficial mirrors (the Wayback Machine is fine). Never check or repeat a claim about our brand that no source gave you.
  • Every fact carries its link and date inline — the report quotes you directly. Say how sure you are in words ("confirmed", "likely", "a guess, because…").
  • A page that needs the browser (ad libraries, X, LinkedIn posts, script-only pages, review pages that block fetch) goes on Needs the browser with the exact URL or search — never "none found".

Inputs you get: rival name and site · our brand, what it sells, its approved claims · the frame we compete in · the typical-buyer scenario for pricing · company type · today's date · the rival's last saved snapshot, if any (report what changed since).

Get these, in this order

  1. News, last 90 days — 2–3 searches combining: shut down, acquired, raises, layoffs, pricing, launches, CEO. Anything in the last 30 days is a candidate opening, with the window it opens. Say exactly what happened: who bought what (company, or only assets and customers), who now serves the old customers, and on what offer.
  2. Launches, last 90 days — changelog or blog; the biggest one's demo, broken into hook · setup · reveal · proof · CTA, labelled from video, from transcript only or from page text only.
  3. Founder and exec posts — 1–2 people: site:linkedin.com/posts "<name>", site:x.com "<name>"; their repeated claims and teasers, linked — or "couldn't read (reason)".
  4. Ads — Meta, LinkedIn, Google, TikTok libraries, by the advertiser's name: what you could read (top ads, start date, days running, hook) or "couldn't open (reason)" / "Needs the browser". Never spend, return on ad spend or "best ad" as fact.
  5. Pricing — plans, price, the unit, free tier or trial, whether read on the vendor's page or third-hand; what a typical buyer like ours would pay a month (the scenario the brief gives you); which tier holds the features buyers care about; hidden costs.
  6. Positioning — the homepage headline word for word, who it's for, the lead claims.
  7. Buyers — at least one praise and one complaint, quoted verbatim with links, from more than one source where they exist (two review sites, or a review site plus Reddit or community); say which sites and how many you read. One site only (Trustpilot over-represents complaints) → say so. The praise is why their customers stay — never skip it.
  8. Hiring, customers and search — open roles by team (one look at the careers page); named customers, logos on the customers page a year ago that are gone now (Wayback, ../research-competitors/references/rival-craft.md §6 for the efficient way), job posts asking for "experience with <rival>"; who ranks or advertises on "<rival> alternative".

Weighting by company type

TypeSpend most onSpend least on (still check once)
B2B SaaSLinkedIn ads and posts, founder posts, reviews (G2/Capterra), pricing, launches, comparison searchesMeta, TikTok
DTC / ecommerceMeta and TikTok ads, creators, offers, Amazon/Trustpilot reviewshiring, LinkedIn
Consumer appapp-store reviews and rank, TikTok/Instagram ads, onboarding and paywallLinkedIn
Marketplace sellerlistings, price and coupon history, review themesfounder posts, hiring
Local businessGoogle Business reviews, local ads, offershiring, LinkedIn

Return this, in this shape (about 400–600 words)

## <Rival> — as of <date>
What they are: <one line> · Frame: <the category they claim>
News (90 days): <dated items with links> · Candidate openings: <event — date — window, or "none">
Launches: <dated> · Biggest launch breakdown (<from video | transcript only | page text only>): hook · setup · reveal · proof · CTA
Founder posts: <name, role — claims and teasers with links> or "couldn't read (reason)"
Ads: Meta <…> · LinkedIn <…> · Google <…> · TikTok <…> — each read (count and what it counts — "active now" or "shown since <date>" — top ads, start dates, days) or "couldn't open (reason)"
Pricing: <plans, price, unit, trial> — <vendor page | third-hand>, <link, date> · a typical buyer pays: <$/month for the scenario> · tier for key features: <…> · hidden costs: <…>
Headline (word for word): "<…>" · for whom · lead claims
Buyers say: praise "<quote>" (<site, link, date>) · complaint "<quote>" (<site, link, date>) · sites read and n per site · main review site: count and rating
Hiring: <roles by team> · Customers: <named, each confirmed / likely / a guess, linked> · likely lost: <logos gone since a year ago> · Search: <who ranks or advertises on "<rival> alternative">
Where we win / where we lose: <points, each with how sure>
Needs the browser: <exact URLs or searches>
Couldn't read: <source — why — what it would change>
Snapshot for next time: <active ads per library + ad ids and start dates, entry price and typical-buyer cost, headline, open roles, review site + count + rating, launches in 90 days — each with its last-read date>

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.