Skip to content

SKILL.md

Latest commit

 

History

History
148 lines (122 loc) · 7.42 KB

SKILL.md

File metadata and controls

148 lines (122 loc) · 7.42 KB
Raw
name linkedin
description When a workflow needs public LinkedIn data through UnifAPI — company pages, follower/employee counts, open jobs and job-count, member insights, people, posts, or engagement. Also use on "research this company on LinkedIn," "who works at," "open roles at," "LinkedIn profile for," "company posts," or when another skill (account research, news signal, buying signal, competitor profiling) needs the deterministic LinkedIn read path. Connect via the `unifapi` skill first. Read-only research, never connects or messages.
license MIT
metadata
author version homepage source
UnifAPI
1.0.0

linkedin

The deterministic read path for public LinkedIn data through UnifAPI. This is a Data Skill: it does not run a marketing job on its own — it names the concrete linkedin/... operations, response shapes, and gotchas so any B2B-first workflow (account research, news signals, buying signals, competitor profiling) reads from one known recipe instead of rediscovering the surface each time.

Read-only — eyes, not hands. It researches public LinkedIn data and returns cited records; it never connects, messages, or applies, and UnifAPI never holds LinkedIn credentials.

Use the unifapi skill for live evidence

Connect once through the shared unifapi skill (OAuth MCP), then call the operations below. Companies are keyed by their public {slug} (the vanity segment of the company URL) and people by their public {username} — read both from the LinkedIn URL, not a numeric id. Keep any billing metadata so the output can state record cost.

Response contract

Single-entity endpoints return the object in data:

{
  "request_id": "unif_...",
  "data": {},
  "billing": { "records_charged": 1, "balance_remaining": 99 }
}

List endpoints return an array in data plus pagination:

{
  "request_id": "unif_...",
  "data": [],
  "pagination": { "has_more": false, "next_cursor": null },
  "billing": { "records_charged": 1 }
}

When pagination.has_more is true, pass pagination.next_cursor as the next request's cursor. Always preserve billing when reporting cost.

Core operations

Need Operation
Company page linkedin/companies/{slug}
Company headcount signal linkedin/companies/{slug}/job-count · .../jobs
Company people linkedin/companies/{slug}/people
Member insights linkedin/companies/{slug}/member-insights
Company posts linkedin/companies/{slug}/posts
Person profile linkedin/users/{username} · .../about · .../experience
Person reach linkedin/users/{username}/follower-count
Person posts / reactions linkedin/users/{username}/posts · .../reactions
Search people linkedin/search/people (?title=&current_company=&industry=)
Search jobs / posts linkedin/search/jobs · linkedin/search/posts
Job / post by id linkedin/jobs/{id} · linkedin/posts/{id} (.../comments)
Resolve a geocode / industry linkedin/search/locations · linkedin/search/industries

Need a field not listed here? Use the unifapi skill's get_operation to read the exact schema before calling — but pick the operation from this table, don't discover blind.

Workflow

The deterministic recipes. Pick the one that matches the job; each names exactly what to call.

  1. Resolve an account. Take the company {slug} from its LinkedIn URL, then call linkedin/companies/{slug} for follower_count, employee_count, industries, and headquarters.
  2. Size hiring as an investment signal. Call linkedin/companies/{slug}/job-count (returns a single total) and linkedin/companies/{slug}/jobs for the open roles — a rising count or a cluster of senior roles is a growth/priority signal.
  3. Map the buying committee. Call linkedin/companies/{slug}/people and linkedin/companies/{slug}/member-insights, or narrow with linkedin/search/people?current_company=...&title=... for specific roles.
  4. Read a person. Call linkedin/users/{username} plus .../about and .../experience; .../follower-count for reach (a separate LinkedinFollowerStats object); .../posts for what they publish.
  5. Read posts and engagement. Call linkedin/companies/{slug}/posts or linkedin/users/{username}/posts; each LinkedinPost carries like_count, comment_count, and share_count. Page via next_cursor.
  6. Search the surface. Use linkedin/search/people|jobs|posts with filters; resolve a geocode_location via linkedin/search/locations and an industry id via linkedin/search/industries first.
  7. Cite everything. Every claim ties back to the company, person, or post it came from; report billing.records_charged (or estimate when billing metadata is absent).

Shape notes

  • LinkedinCompany — keyed by {slug}. follower_count, employee_count, employee_count_range, industries, headquarters, is_verified.
  • LinkedinUser — keyed by {username}. Profile flags at top level: is_open_to_work, is_hiring, is_top_voice, is_creator, is_premium. Follower/connection counts are not here — read them from .../follower-count (LinkedinFollowerStats: follower_count, connection_count).
  • LinkedinPost — like_count, comment_count, share_count, reactions, author, post_type.
  • LinkedinJob — title, location, salary, level, employment_type, listed_at, company. LinkedinJobCount is just { total }.

Gotchas

  • Companies are keyed by {slug}, people by {username} — both read from the public LinkedIn URL, never a numeric id.
  • Follower and connection counts come from linkedin/users/{username}/follower-count, not the base profile object.
  • linkedin/search/people needs at least one filter — there is no all-of-LinkedIn dump.
  • A low balance can silently truncate list pages: check billing.truncated_due_to_balance — when true the page is partial, so top up before trusting any count computed from it.

Output

Return the records the calling workflow needs, each cited to its company, person, or post, plus a one-line cost note (records_charged). When this skill is used directly, a compact account brief is the default:

**{Company}** — {followers} followers, {employees} employees, {industry}. Open roles: {N} ({trend}). Recent posts: {engagement}. Likely buyers: {names/titles}. Evidence: {URLs}. Records: ~{N}.

Related skills

  • linkedin-account-research, account-news-signals (Lead & Company Research) — turn this read path into account briefs and news-tied signals.
  • buying-signal-monitor (Social Selling), competitor-profiling (Competitive Intelligence) — B2B intent and competitor work on top of LinkedIn reads.
  • unifapi — the shared data skill: connect MCP and look up exact schemas with get_operation.