Competitive intelligence your agent can read — correctly
Seven open-source skills that read your CompetLab monitoring data — six dimensions, a Strategic Briefing over fourteen analysis areas — and shape it into what you need: a briefing, a dossier, a battlecard, a work list. The skills do not do the research. The platform does; the skills make sure what comes out the other end is true.
Seven skills, one question each
Every skill answers one question and reads the platform to answer it. Say the phrase, or something like it.
AI Visibility
Which companies do AI models recommend in your category — and are you one of them?
Reads the AI Visibility market map across ChatGPT, Claude, Gemini, Perplexity and Google AI Overviews and answers core or tail before it shows a figure. Presence, ranges and per-engine splits are the mechanism that decides membership, not the answer.
AI Sources
Which pages do the engines read before answering, and who is named on them?
For the two engines that hand back the pages they retrieved — Perplexity and Google AI Overviews — it reports, per engine, which companies each answer named, which pages the engine pulled, and which of those pages name your competitors and not you. That last list is the work list.
Briefing
What changed, what it means, what to do — at pulse, landscape or dimension depth.
Reads the Strategic Briefing the platform already wrote — fourteen analysis areas in numbered editions — at the depth the question needs: a short pulse, the full landscape, or one dimension in depth. Compares editions to say what moved since last time. It reads the briefing; it does not rebuild it.
Competitor Dive
Everything the platform holds on one rival, across six dimensions.
One competitor, every dimension filtered to them, plus the briefing's own researched read on that rival — funding, hiring, launches, reliability, customer voice. A dossier a founder or a sales leader acts on, not a card to scan mid-call.
Battlecard
A 60-second sales reference for a live call.
Why you win, where they are genuinely strong, objection handling, killer facts and landmines — scannable, spoken, dated. Every figure carries the count it came from, so it survives the prospect's follow-up question.
Site Audit
no project neededWhat can a machine actually read on this site? No project needed.
Crawler access, sitemap coverage, agent adoption, tech stack, trust signals — and whether a claimed MCP server answers the protocol or only a browser. Runs on any public domain with nothing set up: no project, no competitors configured. A free-trial key is enough.
Monitoring Setup
Are you watching the right competitors and asking the right questions?
Reviews the roster, the AI Visibility prompts and the per-dimension schedules, and recommends what to change — roster changes as swaps, with the evidence on both sides. It carries no write tools and changes nothing itself.
Try it on any domain
competlab-site-audit runs on any public domain with nothing set up — no project, no monitored competitors. A free-trial key is enough. Crawler access, sitemap coverage, agent adoption, tech stack, trust signals.
The reporting discipline ships with the skills
An agent given competitive data will happily round, pool and rank it into something confident and wrong. Every skill carries the reporting rules the platform enforces on itself — the file that stops it.
null means not measured
Never zero, never empty, never “no”. A real 0 is a finding; a null is a gap in the reading, and the two never share a sentence.
Counts, never rates
“Named in 8 of 69 answers”, not “12%”. The question set is small by design, and a share computed from it is false precision.
Per engine, never pooled
The engines read different pages. A combined figure describes a list none of them produced.
Retrieved, never cited
The engines do not disclose which pages they leaned on, so a citation count is a number that does not exist.
Overlapping ranges are not ordered
Two brands whose ranges overlap are tied, and a difference between two overlapping readings is two readings, not a movement.
Six dimensions the skills read
AI Visibility
Which companies AI models recommend in your category, across ChatGPT, Claude, Gemini, Perplexity and Google AI Overviews
AI Sources
The pages Perplexity and Google AI Overviews read while answering your buyers' questions, and whether you are named on them
Positioning
Homepage messaging: headline, value proposition, CTAs, audience, differentiator
Pricing Intelligence
Plans, billing models, free tiers, enterprise pricing
Content Intelligence
Sitemap analysis, categories, publishing activity, gaps
Tech & Trust Profile
Tech stack, security headers, trust signals, AI crawler access
Beyond the six, the Strategic Briefing researches eight more areas for the briefing alone — fourteen in all. The briefing skill reads them; there is no skill per area because none is needed.
What you get
Example output from competlab-briefing at pulse depth — a single run.
Competitive pulse
Edition No. 07 · read at pulse depth
TL;DR
Nine companies make up this market as the models draw it. You are not one of them — named in 8 of 69 answers across five engines. Competitor X moved its Pro plan from $149 to $119. Competitor Y published 8 articles this cycle to your 2.
Key changes
● Pricing: Competitor X cut Pro from $149 to $119
● AI Sources: Perplexity retrieved your pricing page in 2 of 8 answers and named only other companies
● Content: Competitor Y published 8 articles this cycle, against your 2
Recommended actions
1. Pricing: answer the Pro cut or state the value gap on the page — decide this week
2. AI Sources: two of the retrieved hosts name three competitors and not you — the work list is in the AI Sources section
3. Content: 2 against 8 — close the gap on the buying questions first, not on volume
Give your AI agent competitive intelligence
Start free. Your API key is available immediately, and the skills install in a minute.