market-search
Skill1 starsChanged 9 days agoMITView on GitHub
Broad people, company, and jobs market search via LeadMagic MCP with plan-aware cost, fair-use limits, and cursor pagination. Use when building audiences, TAMs, account lists, or job-posting sweeps rather than single lookups.
In this repository9 other files it gives its agents.
- account-researchSkill
- bulk-enrichmentSkill
- contact-lookupSkill
- credit-guardSkill
- decision-makersSkill
- get-startedSkill
- hiring-intentSkill
- outbound-recipesSkill
- sheetsSkill
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 2 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 9
Use when the user wants a whole segment — an audience, a TAM, an account list, or a job-posting sweep — not a single contact or company …
- Very specific: line 54
Only limit may change between pages.
- Distinctive: line 56
When the response spells out the exact next-page arguments, send exactly those.
- Very specific: line 58
Stop: when has_more is false or next_cursor is null.
≈ “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.
Files
Market search (people · companies · jobs)
Trigger
Use when the user wants a whole segment — an audience, a TAM, an account list, or a job-posting sweep — not a single contact or company lookup.
What costs credits
Browse search means the three canonical searches: search_people, search_companies, and search_jobs.
- On qualifying plans, browse search is credit-free, within two limits:
- Rate limit, set by the plan.
- Fair-use cap: about 10M records a month — 13,333 searches a day at 25-row pages, or your contract's limit if lower. The cap is per account, shared across people, company, and jobs search.
- On other plans, browse search is metered by results returned.
- Always metered, on every plan: lookalikes (
find_lookalike_companies), full search, exports, and contact unlocks (emails, phones).
Which plans qualify, and their rate limits, are on leadmagic.io/pricing. Never promise that search has no limit. If you don't know the user's plan, call the free check_credit_balance and preview_cost before a large pull, and state the expected cost.
Even when search itself is free:
- Size the audience first.
- Unlock contact details last, and only on the final list.
Workflow
- Shape filters with a small probe page (
limit5–10) and adjust them before paging. Tool filters:search_people:company_domain, orcompany_filters(industries,min_employees/max_employees,country_codes,company_domains), plustitles/rolesandpeople_filters.search_companies:company_filtersorquery.search_jobs:titles,company_domains(up to 25),countries,seniority,posted_within,tags,has_remote.
- Run the broad search and page it with cursors (see below).
- Respect the snapshot cap. One browse snapshot returns at most 1,000 rows. For a bigger market, split it into filter segments (industry × size band × country) and page each one separately. Do not try to page past the cap.
- Only then enrich. Get emails per the
decision-makers/contact-lookupskills, or through thebulk-enrichment/sheetsskills for more than 5 people.- Finder emails come back already validated.
- Validate only emails the user brought from elsewhere.
Page size
| Lane | Rows per page |
| Credit-free browse lane | Capped at 25, whatever limit you send |
Metered: search_people | limit up to 100; cursor pages up to 50 |
Metered: search_companies, search_jobs | limit up to 50 |
Send the page size you want. Expect 25-row pages on the credit-free lane.
Cursor pagination (all three searches)
- First page: filters plus
limit. Omitcursor. - Each response carries
next_cursor(opaque) andhas_more. - Next page: the same arguments plus
cursor: <next_cursor>, copied verbatim.- Only
limitmay change between pages. - Changing any other argument invalidates the cursor.
- When the response spells out the exact next-page arguments, send exactly those.
- Only
- Never combine
cursorwith a nonzerooffset. To switch from offset paging to cursors, restart at offset 0. - Stop when
has_moreis false ornext_cursoris null. - Dedupe on stable IDs.
Output
- The paged result set (or the file it was written to), the row count, and the pages fetched
- The filters used, and any part of the segment the filters could not express
- The next enrichment step and its credit cost (from
preview_cost)
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
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.

