agentleFS
elmohqelmohq/elmo/.claude/skills/add-competitor/SKILL.md

add-competitor

Skill577 starsChanged 20 days agoMITView on GitHub

Add a new competitor to the AI Visibility Tool Directory — researches the tool, generates data, takes a screenshot, and inserts into the codebase

  • Reads credentials

The file

We split this skill into the rules it gives an agent

11 rules. About 0 are common, 7 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈7

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Add a new competitor to the AI visibility tool directory at apps/www/src/lib/competitors/data.ts.

    line 9
  2. Very specific:

    Use WebFetch to read the competitor's website at $ARGUMENTS.

    line 17
  3. Very specific:

    Read the feature definitions in apps/www/src/lib/competitors/types.ts (the FEATURE_CATEGORIES constant).

    line 30
  4. Distinctive:

    Create a URL-friendly slug from the product name (lowercase, hyphens, no special chars).

    line 74
  5. Distinctive:

    Run the screenshot script:

    line 78
  6. Distinctive:

    If the script fails, inform the user and continue with the data entry — the screenshot can be added later.

    line 85
  7. Very specific:

    Read apps/www/src/lib/competitors/data.ts and insert the new Competitor object into the competitors array (position in this array doesn't …

    line 97
  8. Distinctive:

    Use the same code style as the existing entries.

    line 99
  9. Distinctive:

    Look at the tools already in each tier to calibrate where the new tool belongs.

    line 148
  10. Distinctive:

    Run TypeScript type checking to make sure the entry compiles:

    line 152
  11. Distinctive:

    After completing all steps, tell the user:

    line 159

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Add a new competitor to the AI visibility tool directory at apps/www/src/lib/competitors/data.ts.

The user will provide a URL. Follow these steps:

Important: Write objective, factual descriptions. Do not use marketing language, exaggerations, or the competitor's own promotional claims at face value. Describe what the tool does and how it works in neutral terms. Avoid superlatives like "industry-leading", "best-in-class", "revolutionary", etc. Stick to verifiable facts — features offered, platforms tracked, pricing tiers. Highlights should be factual differentiators, not praise.

1. Research the competitor

Use WebFetch to read the competitor's website at $ARGUMENTS. Gather:

  • Product name
  • Domain
  • What the tool does (tagline + longer description)
  • Which category it fits: tracking, content, api-developer, ecommerce, seo-traditional, open-source, or other
  • Pricing info (free tier? starting price? enterprise?)
  • Status: active, beta, acquired, shutting-down, defunct, or pivoted (set fate for the last four so the tool shows up correctly in the AEO Graveyard)
  • License (SPDX id, e.g. MIT, AGPL-3.0) if the code is public
  • checkedAt: today's date (YYYY-MM-DD). Comparison pages show it as "last checked", so bump it on an existing entry whenever you re-verify it, even if nothing else changed
  • 3 highlights (short bullet points about what makes it notable)

2. Determine feature flags

Read the feature definitions in apps/www/src/lib/competitors/types.ts (the FEATURE_CATEGORIES constant). Based on your research, determine which features the competitor supports. Set each to true or omit it (defaults to false). The features are:

  • multiLlmTracking — tracks across multiple AI platforms (ChatGPT, Claude, Gemini, etc.)
  • Engines tracked, one flag each: chatgpt, googleAiOverviews, googleAiMode, gemini, perplexity, copilot, claude, grok, deepseek, metaAi — check the pricing page too, since engines are often gated by plan (still true)
  • anyModel — users can add arbitrary models (e.g. any OpenRouter model) rather than a fixed list
  • uiCapture — the vendor says it captures answers from the real consumer apps, not only model APIs
  • byoKeys — customers can plug in their own scraper or model API keys
  • visibilityScore — provides an aggregate AI visibility score
  • citationAnalytics — tracks citation sources in AI responses
  • competitorBenchmarking — compare visibility against competitors
  • brandMentionTracking — monitor brand mentions in AI responses
  • promptVolumeEstimates — estimated search/prompt volumes
  • sentimentAnalysis — brand sentiment in AI responses
  • crawlerAnalytics — track AI bot visits to your site
  • geographicTracking — visibility by region/country
  • socialMediaTracking — Reddit and social platform monitoring
  • shoppingTracking — product visibility in AI shopping
  • multiLanguage — multi-language support
  • actionRecommendations — prioritized action items
  • contentGapAnalysis — detect content gaps vs competitors
  • siteAudits — AI site readiness audits
  • keywordResearch — AI keyword/prompt discovery
  • emailAlerts — automated alerts
  • dataExportApi — CSV export or API access
  • typescriptSdk / pythonSdk — an official, published client library for their API (not just curl snippets)
  • mcpServer — an MCP server for AI assistants
  • biConnectors — Looker Studio, NinjaCat, etc.
  • whiteLabelAgency — white-label or agency features
  • openSource — source code available
  • contentGeneration — AI content creation

Be conservative — only mark features as true if you can confirm them from the website.

3. Fetch the domain rating

Extract the domain from the URL (e.g. https://www.example.com/foo → example.com). Then run:

cd apps/www && node scripts/fetch-domain-rating.mjs "<domain>"

This returns JSON with ahrefsDR, from Ahrefs' free public API (no API key required). Use it for the competitor entry.

4. Generate the slug

Create a URL-friendly slug from the product name (lowercase, hyphens, no special chars). For example: "Otterly.ai" → "otterly", "SE Ranking" → "se-ranking".

5. Take the screenshot and upload

Run the screenshot script:

cd apps/www && node scripts/screenshot-competitor.mjs "<slug>" "<url>"

This requires SCREENSHOT_ONE_ACCESS_KEY and BLOB_READ_WRITE_TOKEN in apps/www/.env.

If the script fails, inform the user and continue with the data entry — the screenshot can be added later.

Verify the capture — don't trust it blindly. ScreenshotOne uploads whatever the site serves, and sites with bot protection (Cloudflare, etc.) frequently return a captcha or block page to the headless browser, which then gets stored as the "screenshot." After a successful upload, download the blob URL the script printed and actually view the image:

curl -s -o /tmp/<slug>.jpg "<blob URL printed by the script>"

Open /tmp/<slug>.jpg and confirm it's genuinely the competitor's homepage/product — not a bot challenge ("Checking your browser…", "Select all squares with…", "verify you are human"), an access-denied / 403 page, or a blank/error page. If it's a challenge or error page, do NOT keep it: tell the user automated capture was blocked and ask them to send a real screenshot to upload manually (the script sets allowOverwrite: true, so re-uploading replaces the bad image).

6. Insert into data.ts

Read apps/www/src/lib/competitors/data.ts and insert the new Competitor object into the competitors array (position in this array doesn't matter for display order).

Use the same code style as the existing entries. Example entry:

{
    slug: "example-tool",
    name: "Example Tool",
    domain: "example.com",
    url: "https://example.com",
    tagline: "Short one-line description of what they do",
    description:
        "Longer 2-3 sentence description of the tool, its approach, and what makes it notable.",
    category: "tracking",
    ahrefsDR: 55,
    status: "active",
    checkedAt: "2026-11-03",
    features: {
        multiLlmTracking: true,
        visibilityScore: true,
        citationAnalytics: true,
    },
    pricing: { hasFree: true, startingPrice: "$49/mo", hasEnterprise: false },
    highlights: [
        "First highlight",
        "Second highlight",
        "Third highlight",
    ],
},

7. Add to AEO popularity ranking

In the same file (data.ts), find the aeoPopularityRanking array. This is a hardcoded list of slugs ordered from most popular to least popular as an AEO tool specifically. It determines both the display order in the directory and the A–F popularity grade.

Insert the new tool's slug at the appropriate position based on these criteria (in priority order):

  1. Is AI visibility / AEO the tool's PRIMARY product? Pure AEO trackers rank above SEO suites or content platforms that bolted on AEO features.
  2. Market traction as an AEO tool — funding rounds, notable enterprise customers, press coverage (Fortune, WSJ, TechCrunch, etc.), user counts, industry recognition (G2, Gartner, etc.).
  3. Feature depth in core AEO capabilities — multi-LLM tracking, visibility scores, citation analytics, competitor benchmarking, brand mention tracking.
  4. General web authority (DR/traffic) is a minor tiebreaker only — a DR 90 SEO suite with a small AEO add-on should NOT outrank a DR 50 dedicated AEO tracker.

The tiers in the ranking are:

  • Tier A (top ~3%): Category leaders with major funding, customers, and recognition
  • Tier B (next ~7%): Well-funded or widely adopted AEO tools
  • Tier C (next ~15%): Known/growing AEO tools
  • Tier D (next ~25%): Niche/emerging AEO tools with some traction
  • Tier F (rest): Early, small, or minimal-traction tools

If the tool is a large platform where AI visibility is NOT the primary product (e.g., a traditional SEO suite, analytics platform, or content tool that added AEO as a secondary feature), do NOT add it to aeoPopularityRanking. Instead, add its slug to the aeoNotApplicable set. These tools get an "N/A" grade and sort to the bottom of the directory.

Look at the tools already in each tier to calibrate where the new tool belongs. Insert the slug with a short comment explaining the placement.

8. Verify

Run TypeScript type checking to make sure the entry compiles:

pnpm exec tsc --noEmit

9. Summary

After completing all steps, tell the user:

  • The competitor name and slug
  • The category and ahrefsDR
  • Which features you marked as true
  • Whether the screenshot uploaded successfully
  • The URL where the comparison page will be: /ai-visibility-tools/elmo-vs-<slug>

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.