agentsclimarketplace

Segment intelligence

Skill intelligems-io/intelligems-mcp-skills/skills/segment-intelligence

Ready-to-install Agent Skills for analyzing Intelligems experiments, segments, pricing, and reporting through MCP and API.

Install
npx -y skills add intelligems-io/intelligems-mcp-skills --skill segment-intelligence

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

Use when a user asks how Intelligems test performance differs by device, visitor type, country, traffic source, source site, or landing page.

The file declares its own license as MIT. 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

1.7 KB, 334 tokens by cl100k_base, as published. Nobody here has run it

Segment Intelligence

Use this skill to identify customer segments that consistently overperform or underperform.

Segment Tools

Current public MCP docs include:

  • get_audience_by_country
  • get_audience_by_device
  • get_audience_by_visitor_type
  • get_audience_by_source_channel
  • get_audience_by_source_site
  • get_audience_by_landing_page

You can also use analyze_experience with view: "audience" and the appropriate audience field when the active client exposes that path.

Steps

  1. Resolve organization, experience scope, and date range.
  2. Identify relevant tests with search_experiences, search_experiments, or explicit experience IDs from the user.
  3. Pull overview metrics first so segment findings have context.
  4. Pull only the segment dimensions needed for the user question.
  5. Compare segment behavior across tests to separate one-off noise from repeated patterns.

Output

Return:

  1. Strongest segment insights.
  2. Segment table with metric direction, confidence, and sample caveats.
  3. Segment-specific recommendations.
  4. Tests to run next if the pattern is actionable.
  5. Appendix with dimensions queried, tool calls, and any missing metrics.

Do not overstate tiny segments. If sample size is thin, label the finding as directional.

Data Safety

Use only the user's authenticated MCP or API context. Do not save raw responses, exports, graph URLs, order data, customer data, or private store identifiers to the repository. Summarize findings and include experience IDs only when needed for the user's workflow.

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.