Analyse competitors and alternatives: positioning, pricing, offers, which channels and content formats work for them, top hooks, ads they run, and the gaps we can own; writes competitors.md. Use for \"analyse competitors\", \"what is X posting\", \"competitor content audit\", \"ad library research\", \"find gaps\". Not for customer VoC (customer-research) or daily trend tracking (trend-radar).
Scanned 9/28/2026
Install to Claude Code
npx -y skills add brunopass/mkt-harness --skill competitor-intel --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Competitor Intel?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/brunopass-competitor-intel)More formats (shields.io, HTML) on the badges page.
---
name: competitor-intel
description: "Analyse competitors and alternatives: positioning, pricing, offers, which channels and content formats work for them, top hooks, ads they run, and the gaps we can own; writes competitors.md. Use for \"analyse competitors\", \"what is X posting\", \"competitor content audit\", \"ad library research\", \"find gaps\". Not for customer VoC (customer-research) or daily trend tracking (trend-radar)."
---
# competitor-intel
Know what the alternatives say, what works for them, and where they leave room for us.
Global rules: [AGENTS.md](../../AGENTS.md). Research is read-only: never follow, like, comment or DM from a brand
account while researching.
## Inputs
- `brand`; competitor list from `brand.md` (Unlike…), `competitors.md`, customer-research ("we already use X"),
or ask for 3-6 names. Include 1-2 indirect alternatives (spreadsheet, agency, freelancer, doing nothing).
## Workflow
1. `brand_context {brand}`; read `competitors.md` and `insight_list {brand, kind: ["competitor"]}`.
2. **Website pass**: `site_scan {url, brand}` gives a structured first read (positioning line, languages, socials,
contacts, look) saved to `research/<host>.md`; then WebFetch for depth: home headline + subhead, category they claim, pricing page (tiers, anchor price,
free trial/freemium, annual discount), guarantee, main CTA, proof they show. Capture the headline verbatim.
3. **Social pass** (browser, read-only, on a logged-in brand account of the same platform):
`browser_open {account, url: "<competitor profile>"}`, `browser_scroll`, `browser_text`, `browser_screenshot`
when a visual pattern matters. For each platform they're active on, record for the last 30 posts:
format (reel/carousel/text/video), hook (first line / first 2s on-screen text), topic/pillar, CTA, and visible
engagement (views/likes/comments). Rank by engagement relative to their follower count; study the top 5 and the
bottom 5 (what to avoid).
4. **Ads**: Meta Ad Library (`https://www.facebook.com/ads/library/` — country + name), TikTok Creative Center
(Top Ads, filter by industry/region), LinkedIn company page → "Ads" tab / LinkedIn Ad Library, Google Ads
Transparency Center. Long-running ads (> 30 days) are the ones that convert — note their angle, offer and format.
5. **Customers' view**: 3★ reviews of each competitor (what they fail at) → `insight_add` with `kind: "competitor"`
and verbatim quotes.
6. **Synthesis** → write `workspace/brands/<brand>/competitors.md`:
- Table (template columns): Competitor | Positioning (their headline) | Price | Channels that work |
Top formats/hooks | Gaps we can own.
- Per competitor: 3-5 bullet teardown; top 3 posts with URL + why it worked (hook type, emotion, format, timing).
- **Pattern section**: hooks/formats that work across ≥ 2 competitors (table stakes), and angles nobody takes.
- **Gaps**: underserved persona, unanswered objection, unowned format (e.g. nobody does founder-led video in pt-BR),
price/packaging gap, channel gap. Each gap → one concrete content or offer idea.
7. Optionally create 3-5 `content_create {brand, title, status: "idea", hook, pillar, persona, body}` ideas from the
gaps (idea-engine will score them).
8. Report the 5 most important findings, each with evidence.
## Metrics discipline
- Engagement rate proxy = (likes + comments) / followers, or views / followers for video. Note the date seen.
- Don't compare raw numbers across platforms. Compare within a platform and time window.
- Numbers you couldn't see → leave blank, don't estimate.
## Quality bar
- Every claim about a competitor links to where you saw it (URL + date).
- At least 3 gaps, each tied to evidence (a review, a missing format, an unanswered objection).
- Clear "steal / avoid / own" recommendation per pattern.
## Pitfalls
- Copying competitor hooks word-for-word: extract the structure, rewrite in our voice with our proof.
- Assuming high follower count = what works now. Look at recent posts and long-running ads.
- Logging into or interacting from the brand account (visible to them, and can look like spam).
## Show it
When done, show `competitors.md` as a comparison page (positioning, pricing, channels, gaps): an Artifact in Claude Code, else `report_open {path}`. See AGENTS.md › Show the result.
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!