agentsclimarketplace

Vantage discover

Skill macmenes/vantage/skills/vantage-discover

Vantage.os — Marketing Intelligence OS. The marketing brain inside your agent. /vantage interviews you, then delivers a beautiful, evidence-backed competitive teardown (Share-of-Voice, positioning map, battlecards). Web-only, no API keys.

Install
npx -y skills add macmenes/vantage --skill vantage-discover

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

TRIGGER when the user needs to identify and tier the real competitor set for a brand/domain — "who are our competitors", "find our competitors", "build a competitor list", "competitive set", "who else is in this market". Part of the `vantage` suite. Produces a tiered, evidence-backed competitor set. DO NOT trigger for analyzing competitors already named (use the relevant vantage-* channel skill) or keyword-ranking rivals only (use seo-dataforseo).

SKILL.md

3.6 KB, as published. Nobody here has run it

Vantage-Discover — Build the Real Competitor Set

The most common failure in competitive analysis is analyzing the wrong set — the competitors leadership names instead of the ones the buyer actually considers. This skill builds the set from market evidence, tiers it, and caps the deep-dive to a workable size.

Web-only. See ../vantage/references/data-sources.md for sourcing patterns.

Method

  1. Anchor on the subject brand. WebFetch its homepage; extract category, ICP, and the buyer problem it solves. The set is "who else solves this for this buyer."
  2. Cast the net across five lenses (each surfaces a different rival type):
    • SERP rivals — run 5–10 buyer queries for the category; collect domains that repeat on page 1. These are who the buyer sees.
    • "Alternatives / vs" mining"<brand>" alternatives, "<brand>" vs, best <category> 2026. Read the listicles and G2 "compare" tabs.
    • Review-site neighbors — on G2/Capterra, the "competitors" and "compare" modules name the considered set directly.
    • Ad rivals — who bids on the brand term (Google Ads Transparency) and who runs ads on the same angles (Meta Ad Library keyword search).
    • Emerging/funded<category> startup funding 2025 2026; catch entrants before they're obvious.
  3. De-dupe and classify each candidate into a tier (below).
  4. Tier and cap. Keep the top 5–7 for deep-dive by relevance to the subject's actual buyer; list the long tail so nothing's lost.

Tiering model

TierDefinitionTreatment
DirectSame buyer, same problem, substitutableFull deep-dive
AdjacentOverlapping buyer or partial-solution overlapDeep-dive if it appears in deals/SERPs often
AspirationalThe category leader / who everyone benchmarksAnalyze for positioning even if "out of league"
EmergingNew/funded entrant changing the rulesWatch; analyze if momentum is real
IndirectThe status quo / DIY / "do nothing" / spreadsheetsName it — often the real competitor for budget

Don't forget the indirect tier. For many growth/sales motions the biggest competitor is "the buyer keeps doing it manually," and the messaging must beat that.

Relevance scoring (to pick the deep-dive set)

Score each candidate 0–3 on: buyer overlap, problem overlap, how often it co-appears (SERP/reviews/deals), and momentum. Sum → rank → take the top 5–7. Show the scores so the cut is defensible.

Output

# Competitor Set — [Subject Brand]
*Category: [category] · Buyer: [ICP] · Prepared [date]*

## Deep-dive set (top N by relevance)
| Competitor | Tier | Buyer overlap | Why in set | Relevance |
|-----------|------|---------------|-----------|-----------|

## Long tail (named, not deep-dived)
[list with one-line each]

## Notable absences / blind spots
[markets/segments we couldn't see; sources that blocked fetching]

## Recommended deep-dive set → hand to /vantage
[the 5–7 to run channel teardowns on]

Then offer: "Run the full /vantage teardown on this set?"

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.