SkillProof tested Claude skills

Analyze Competitors

Dogłębne analizy konkurencji z tagami źródła i pewności, a nie zapamiętane fakty

Autor: arezous · arezous/pm-pilot

Testowano · Działa ★ 8.4/10

Analyze Competitors — Dogłębne analizy konkurencji z tagami źródła i pewności, a nie zapamiętane fakty

Co robi ten skill

Przeprowadza analizę konkurencji w jednym z czterech trybów: single-competitor deep dive, full landscape, synthesis of existing research, or a monthly monitoring update. Uruchamia się na requests like „research Linear”, „who are our competitors”, „synthesize competitors” or „monthly competitor check-in”. Najpierw sprawdza internal notes, fills gaps with web research, tags every key claim with a source i confidence level, i writes a structured report to output/competitors/.

Raport z testu

Uruchomiono deep-dive mode on Linear against a from-memory baseline of the same request: the baseline produced 264 words with zero URLs i zero dated facts, the skill run produced a report with 7 live source URLs, 10 source-and-confidence tags i 13 dated changelog entries. The workflow is what forced the extra evidence: it pushed past the pricing page into the changelog i job board, which turned the baseline's guess „probably moving upmarket” into a specific finding (five staffed GTM roles plus a Compliance and Trust Lead). When G2 returned HTTP 403 the skill's citation rules made the run state the source was unreachable rather than invent review quotes, i it flagged three memory-based claims as „[Assumption, verify]” that the baseline had asserted as fact. One rough edge: the output template lives in the repo root, not in the skill folder, so keep the clone around rather than copying only the skill directory.

Testowano: 2026-07-29 · Claude Code 2.x (agent harness)

Instalacja

git clone https://github.com/arezous/pm-pilot.git
mkdir -p ~/.claude/skills
cd pm-pilot && cp -r .claude/skills/analyze-competitors ~/.claude/skills/analyze-competitors

Komendy i przykładowe prompty

  • /analyze-competitorsDogłębne analizy konkurencji z tagami źródła i pewności, a nie zapamiętane fakty

Skille uruchamiają się na zwykłe polecenia — bez komend do zapamiętania. Po instalacji aktywują go prompty takie jak te (po angielsku):

  • Do a deep dive on Linear as a competitor to our product
  • Who are our competitors in the AI project management space?
  • Run our monthly competitor check-in and tell me what changed

Pełny raport →