LinkedIn Agent
Research public LinkedIn company and member profiles, posts, jobs, and comments for B2B — read-only, inside your assistant.
Using ChatGPT, Claude Code or another client? See the setup steps
Run B2B research on public LinkedIn company and member profiles, posts, comments and job listings — it never connects or messages for you.
LinkedIn Agent composes the Role-Agent Skills that read LinkedIn and exposes its public-data API. Same capabilities, sliced by source — read-only.
A sourced B2B brief from current public LinkedIn profiles, posts, jobs, and comments.
- Company profile and recent public company posts
- Targeted job-search evidence with selected job details
- Known member profiles/posts and inferred buyer functions when relevant
- Each claim sourced to the public LinkedIn record
Who runs this Agent
- B2B sales and recruiting teams researching accounts before outreach
- Account planners reading public company, targeted hiring, and post signals
- Competitive teams profiling a rival from its LinkedIn footprint
Paste this into Claude or Codex
The prompt is editable. Replace the brand, market, competitors, or location, then let the assistant call UnifAPI MCP when it needs live public data.
Research this company on public LinkedIn: pull the company profile and recent posts, run one relevant company-filtered job search, and inspect known member profiles only if supplied. Return a sourced brief with inferences and coverage limits marked — read-only, no connections or messages.
From public records to a decision brief
Resolve the account
Start from a company slug, profile URL, or name; the agent picks the narrowest public LinkedIn reads.
Pull public B2B records
UnifAPI returns the current company/member profile and post records plus targeted job and post-detail reads — joinable with news context — in one envelope.
Write the account brief
The agent reports priorities, targeted hiring evidence, inferred buyer functions, and coverage gaps, then hands off to the composed Skills.
The concrete tasks LinkedIn Agent runs
A Skill is one concrete capability — a run-prompt the Agent executes. Open a Skill for its detailed workflow, expected output, and FAQ.
LinkedIn data
EnhancedThe deterministic guide to LinkedIn's eight supported operations — company and member profiles/posts, jobs, post detail, and comments.
Account research
EnhancedUse public company, post, job, and profile signals for B2B research.
Account news
EnhancedPull recent public news for a target account and convert each event into a dated, sourced outreach hook.
Buying signals
EnhancedWatch public X for buying intent, then qualify named accounts with current LinkedIn and news evidence.
Competitor profiling
EnhancedTurn a competitor's public footprint into a structured, source-cited profile — positioning, product, pricing, and implications.
Public-data tools this Agent calls
Agents are the workflow layer. The same live APIs are callable directly for product code, debugging, and custom agent flows.
This is a read-only research agent for one platform. It reads public data and returns evidence — it never logs into, posts to, comments on, or messages on the platform, and it never holds your account credentials.
Questions about the LinkedIn Agent
Scope, billing, public-data boundaries, and how this Agent runs inside your assistant.
It reads LinkedIn's public surface through the UnifAPI catalog — public profiles, company pages, posts, and jobs, keyed by the URL slug. It does not log into your account or read private data.
No. The LinkedIn Agent is read-only research. You send any outreach from your own account; UnifAPI never holds LinkedIn credentials.