agentsclimarketplace

Mentions discover

Skill nikhonit/llm-mentions-skills/skills/mentions-discover

Agent skills to track brand mentions, ranks & citations across ChatGPT, Claude, Gemini & Perplexity — AI brand monitoring + GEO for Claude, Cursor & any agent

Install
npx -y skills add nikhonit/llm-mentions-skills --skill mentions-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

Discover the queries worth tracking for a brand's AI visibility — informational, commercial, and comparison-intent prompts — via MentionsAPI.com. One tool, minimum surface area.

The file declares its own license as MIT-0. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

3.7 KB, 742 tokens by cl100k_base, as published. Nobody here has run it

mentions-discover

Focused query-discovery skill. Use only when the user explicitly asks what prompts or questions they should be tracking for a brand's AI visibility — the starting query set for a GEO/monitoring program.

When to use this skill

DO use when the user asks:

  • "What queries should I track for Linear?"
  • "Which prompts should I monitor to measure my AI visibility?"
  • "Give me the commercial-intent questions buyers ask in the project-management space"
  • "What comparison queries mention my competitors?"

Do NOT use when:

  • The user already has a specific query and wants to check it — use mentions-check instead
  • A brand is mentioned in passing without a request to build a tracking set
  • The user has not signaled they want query suggestions

This call costs $0.50. Confirm the brand (and industry, if known) before calling.

Tools

discover_queries — $0.50 per call

Return a ranked set of candidate queries to track for a brand, spanning informational, commercial, and comparison intents.

Arguments:

  • brand (required) — the brand to generate tracking queries for, e.g. "Linear"
  • industry — optional context to sharpen suggestions, e.g. "project management software"
  • count — how many queries to return, 1–100 (default 50)

Example response:

{
  "brand": "Linear",
  "candidates": [
    { "query": "best project management tool for engineers", "intent": "commercial" },
    { "query": "Linear vs Jira",                              "intent": "comparison" },
    { "query": "what is Linear used for",                     "intent": "informational" }
  ],
  "cost_cents": 50,
  "balance_after_cents": 50
}

Feed the returned query strings into mentions-check (or mentions-full's check_mentions) to measure visibility, or into watch to monitor them over time.

Authentication

Set MENTIONSAPI_KEY to your MentionsAPI key (format lvk_live_...). Free key with $1 credit at https://mentionsapi.com/signup — no card.

Pricing

Pay-as-you-go. No subscriptions, no monthly tiers, credits never expire. $1 free signup credit, $5 minimum top-up. discover_queries is $0.50 per call regardless of count. Failed calls are not billed.

Errors

Every call returns a Python dict. On success it's the API response; on failure it has an error key (auth_required, auth_invalid, insufficient_balance, rate_limit_exceeded, network). See the full skill for the complete error contract.

API reference

Independence and trademarks

MentionsAPI is an independent service and is not affiliated with, endorsed by, or sponsored by OpenAI, Anthropic, Google, Microsoft, or Perplexity. All product names are the property of their respective owners.

What ships with it: 2 files

6.7 KB alongside SKILL.md, 2 of them executable

tests/

Keep looking

Skills are one crate of 326,984. 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.