Przejdź do treści
Biblioteka skilli
Marketing i treści · Research

competitive-intel

/competitive-intel

Śledzi, co konkurenci mówią, kosztują i robią, na podstawie ich publicznych stron, z zacytowanymi treściami i datami.

Licencja
MIT
Ostatnia zmiana repozytorium
04.11.2026

Ocena i bezpieczeństwo

100/ 100Klasa: dobry
Format25 / 25
Frontmatter z nazwą i opisem, poprawna nazwa, opis do 1024 znaków, nazwa zgodna z folderem.
Odkrywalność25 / 25
Opis mówi, kiedy użyć skilla, ma sensowną długość i jest w 3. osobie.
Treść35 / 35
Co najmniej 150 słów i do 500 linii, nagłówki, listy kroków, przykłady albo kod.
Higiena15 / 15
Bez zaślepek typu TODO, bez ścieżek Windows, z licencją.
Ryzykobrak sygnałów w skanie
Pliki w paczce
3
Skrypty
1
Komentarze HTML
0
Adresy URL w SKILL.md
1
Przestarzałość

Brak sygnałów. Repozytorium zmienione 04.11.2026, bez odwołań do wycofanych modeli i API.

Licencja

MIT, otwarta: można używać i zmieniać, z zachowaniem informacji o autorze.

Co sprawdzić przed instalacją
  • 1 skrypt w paczce. Przejrzyj je przed instalacją, bo Claude może je uruchomić.

Skrypty: scripts/watch_pages.py

Co robi i kiedy

Śledzi, co konkurenci mówią, kosztują i robią, na podstawie ich publicznych stron, z zacytowanymi treściami i datami.

Oryginalny opis (EN)

Know what competitors say, charge and do, from their own public pages and public data, with every claim quoted and dated. Three modes - profile (positioning, pricing, packaging and claims for one competitor), watch (the shared page watcher that flags when a competitor's pricing, docs or claims change, also used by comparison-pages, citation-outreach and entity-presence), and teardown (a dated history of a competitor's growth moves - traffic trend, ads, rankings, content cadence, launches, hiring, reviews - turned into plays for this product). Use when the user asks to profile or watch a compet

Jak działa

  1. State
  2. Profile mode
  3. Watch mode
  4. Teardown mode
  5. Fairness and limits
  6. Spend
Sekcje z nagłówków SKILL.md, w oryginale · 96 linii, 791 słów
Podgląd SKILL.md
---
name: competitive-intel
description: Know what competitors say, charge and do, from their own public pages and public data, with every claim quoted and dated. Three modes - profile (positioning, pricing, packaging and claims for one competitor), watch (the shared page watcher that flags when a competitor's pricing, docs or claims change, also used by comparison-pages, citation-outreach and entity-presence), and teardown (a dated history of a competitor's growth moves - traffic trend, ads, rankings, content cadence, launches, hiring, reviews - turned into plays for this product). Use when the user asks to profile or watch a competitor, "what is X doing", "how did X grow", for a competitive teardown, or when gtm's watch job is due. Comparison pages themselves are ../comparison-pages; the product's own positioning is ../product-context.
---

# Competitive intel

Competitor facts age fast and wrong ones are expensive: a stale price on a comparison page is a
false-advertising risk, and a plan built on a rumour wastes a quarter. So this skill keeps one rule
above the rest: **every statement about a competitor is a quote from a source you read, with its
URL and the date you read it.** Opinions about what it means go in a separate section, labelled.

`<state>` is the store root printed by `python3 ../gtm/scripts/gtm.py where` (field `root`). Run
scripts from the product repo root, calling the plugin's scripts by absolute path (`scripts/…` is
this skill's folder).

## State

```
<state>/competitors/<slug>.md            profile / dossier: sourced facts (shared with ../comparison-pages)
<state>/competitors/<slug>/teardown.md   teardown history and plays
<state>/watch.json                       pages being watched, by purpose
<state>/watch/<id>/, watch/report.md     snapshots and the latest change report
```

Competitors come from `<state>/product-context.md` (alternatives) and from the brands the audit
names instead of you (`named_instead` in the latest `<state>/runs/*/score.json`).

## Profile mode

For one competitor, read their own pages through `treg.web.extract` (header
`X-Treg-Route-Prefer: tinyfish`, free) and record in `<state>/competitors/<slug>.md`:

| section | from | notes |
|---|---|---|
| Positioning | homepage, about page, meta description | their one-liner and category words, quoted |
| Pricing and packaging | pricing page | plans, prices, limits, trial terms; currency and region as shown |
| Claims | product, security and comparison pages | each claim quoted; mark ones that conflict with your claims ledger |
| Who they target | case studies, industries pages, job posts | quote, don't infer |
| Where they show up | ads (below), the AI answers in your audit runs | which prompts name them, which pages cite them |

Every row: the quote, the URL, the date read. Add each source page to the watch list (below) with
`--used-by` naming the pages or dossiers that rely on it. A profile is a snapshot; teardown is the
history.

## Watch mode

The watcher is shared: `../comparison-pages`, `../citation-outreach` and `../entity-presence` use
the same script and file.

```
python3 scripts/watch_pages.py                                   # all watched pages (gtm's `watch` job)
python3 scripts/watch_pages.py --only competitor                 # one purpose group
python3 scripts/watch_pages.py --add https://x.com/pricing --purpose competitor --used-by <page or dossier>
```

| purpose | what a change means | who acts |
|---|---|---|
| `competitor` | their pricing, docs or claims changed | `comparison_refresh` in gtm re-checks the pages listed in `used_by` |
| `target` | a page you pitched started or stopped mentioning you | `../citation-outreach` updates the target's status |
| `profile` | a directory or review profile no longer carries the canonical description | `../entity-presence` |

Changes land in `<state>/watch/report.md`; gtm turns a "Changed" entry into a `comparison_refresh`
job on the next tick. The watcher is free (treg's routed extractor, falling back to a plain fetch).
Record the job with `gtm.py record watch`.

## Teardown mode

A teardown explains how a competitor grew, in date order, from public evidence, and ends in plays
this product can run. Method, sources and the output format: `references/teardown.md`.

1. **Scope.** One competitor, a window (default the last 24 months), and a spend cap (default $2;
   check with `gtm.py budget --need <usd>` first and state the estimate).
2. **Curve.** Monthly traffic estimates for their domain, so you can see when growth happened.
3. **Moves.** For the months around each inflection: ads they ran, keywords they started ranking for,
   content they published, launches, pricing changes, hiring, review volume, press.
4. **Lined up, not proven.** A move that coincides with growth is a candidate, not a cause. Mark each
   play with the evidence strength (several signals agree, one signal, coincidence only).
5. **Plays.** For each candidate worth trying, name the gtm tactic and skill that would run it here
   (`../gtm/references/channel-fit.json`), and whether its channel is eligible for this product
   (`gtm.py fit`). A play on an ineligible channel is noted and not proposed.
6. Write `<state>/competitors/<slug>/teardown.md` and propose the top 3 plays as backlog items in
   the digest. They become bets only through a work session.

## Fairness and limits

- Quote and date every competitor fact. Paraphrase only in your analysis section, which is labelled
  as analysis.
- Traffic, keyword and ad-spend numbers from third-party tools are estimates. Say so next to every
  number, with the provider.
- Never sign up, buy, or contact a competitor's staff while posing as a customer. Read public pages
  only. Never collect personal data about their employees beyond what a job post states.
- Reviews: report patterns with counts, never quote an individual reviewer by name.

## Spend

Profiles and the watcher are free or nearly free. A teardown typically costs $0.50-2.00 (traffic
history and ranked keywords are the expensive calls). Record the spend with `gtm.py record` and
report it in the digest.
Pliki w paczce (3)
  • SKILL.md 6 KB
  • references/teardown.md 4 KB
  • scripts/watch_pages.py 7 KB

Wersje

RepozytoriumWynikLicencja
Ta strona · williampsmith/marketing-skills100MIT
borghei/Claude-Skills100MIT
Hyde-and-Larkin/product-marketing-toolkit100CC-BY-4.0
LeadMagic/gtm-skills100MIT
clairelily2020/b2b-growth-kit100MIT
cyranob/web-forager100MIT
RizwanZafaris/product-manager-OS100MIT
xpozpublic/xpoz-agent-skills100MIT
LeoYeAI/openclaw-master-skills100MIT

Migawka bazy z 13.11.2026 · wersja paczki ecbe5c840b64 (skrót SHA-256 całej paczki)

Pobrania w ostatnich 30 dniach: 0 · ranking

Zgłoś problem z tym skillem: szkodliwa treść, błędny opis albo prośba autora o usunięcie.

Podobne skille

  • 100comparison-pages
    Buduje i prowadzi uczciwe strony typu „X kontra Y”, na które powołują się wyszukiwarki AI, z dossier konkurentów i sprawdzeniem cytatów.
  • 100competitive-pricing-intel
    Monitoruje strony z cennikami konkurentów przez scraping i migawki Web Archive, śledzi zmiany planów i pakietów oraz tworzy macierz porównania cen.
  • 90competitor-profiling
    Tworzy dossier konkurentów na podstawie ich adresów URL i porównuje ich na tle rynku.
  • 90competitor-profiling
    Researchuje i profiluje konkurentów na podstawie listy ich adresów URL.
  • 100competitors
    Tworzy strony porównań i alternatyw konkurentów do SEO i dla kupujących („X kontra Y”, „alternatywa dla X”), a także audytuje istniejące strony konkurencyjne i sprawdza ich aktualność.
  • 90competitor-intel
    Bada konkurentów małych firm: ceny, pozycjonowanie, działania marketingowe i rekomendacje, jak się wyróżnić.

Uwagi i zgłoszenia

Wiadomość trafi do Bartosza Sokolińskiego. Adres e-mail jest potrzebny tylko do odpowiedzi. Szczegóły w polityce prywatności.