CLI

linkedin <command> [args] [flags]

Run linkedin <command> --help for the full flag list on any command. This page is the map. profile, company, job, and jobs work for anonymous visitors, and post is best effort but generally returns data for single public posts and articles. When a page is behind the sign-in wall, linkedin exits with code 4. See troubleshooting.

Commands

Command	What it does
profile	Fetch public member profiles from the Person JSON-LD
company	Fetch company pages from the Organization JSON-LD
job	Fetch a job posting from the guest job-detail fragment
jobs	Search the jobs board through the anonymous guest endpoint
post	Fetch a single public post or article (best effort)
id	Classify and normalize an input into (kind, id) without fetching
url	Build a canonical LinkedIn URL from a kind and an id
db	Inspect the local SQLite record store
cache	Inspect and clear the on-disk page cache
info	Print the resolved configuration
version	Print version, commit, and build date

profile

linkedin profile <slug|url> [slug|url ...] [flags]

Fetches one or more public member profiles, parsed from the page's Person JSON-LD. Accepts a slug (williamhgates), an /in/<slug> path, or a full URL. Profile pages return 200 and work; an exit 4 here usually means IP-level rate-limiting (slow down with --rate or lend --cookies).

Fields include the name, headline, location, country, followers, current and past positions (works_for, alumni_of), member_of (boards and groups, an array of affiliations with name, url, slug, start_date, end_date), and the canonical URL. There is no connection count; LinkedIn does not expose one anonymously.

Flag	Meaning
--posts	Also emit the member's recent posts (the DiscussionForumPosting nodes in the page's JSON-LD @graph) as Post records
--articles	Also emit the member's long-form articles (the Article nodes in the same graph) as Article records
--save	Upsert each profile into the store

If both --posts and --articles are given, --posts wins.

linkedin profile williamhgates --output json
linkedin profile williamhgates --posts

company

linkedin company <slug|url> [slug|url ...] [flags]

Fetches one or more company pages from the Organization JSON-LD plus the company about panel. Fields from the JSON-LD: name, description, website, employees (a point estimate from numberOfEmployees), and the canonical URL. Fields from the about panel and og:description: followers, industry, company_size (the UI band like "10,001+ employees", separate from employees), company_type (for example "Public Company"), founded (the year, when the company lists it; some like Microsoft omit it), specialties (a comma list), and headquarters (for example "Sherman Oaks, CA"). When the company lists funding, the record also carries funding_rounds (the round count) and funding_url (a Crunchbase link to the latest round).

--posts instead collects the company's recent public posts (the DiscussionForumPosting nodes in the main page's JSON-LD graph; the /posts/ subpage is login-walled). --locations emits the full office list, one record per office, with primary marking the registered headquarters and address holding the city, region, postal code, and country line. --affiliated emits the related affiliated and showcase pages, each with its slug, name, industry, and location. --save upserts each company into the store.

linkedin company microsoft --posts
linkedin company microsoft --locations
linkedin company microsoft --affiliated
linkedin company xsolla --fields name,founded,headquarters,followers

job

linkedin job <id|url> [id|url ...]

Fetches one or more job postings from the guest job-detail fragment. Fields include the title, company, company_logo, location, applicant count, posting date, full description, and criteria (seniority, employment type, job function, industries).

linkedin job 3801234567 --output json

jobs

linkedin jobs <keywords...> [flags]

Searches the jobs board through the anonymous guest endpoint, paginating in pages of 25 until -n results or the endpoint runs dry. Emits JobStub records by default (title, company, company_logo, location, id, URL), or full Job records with --hydrate.

Flag	Meaning
--location	Free-text location (e.g. Remote, "United States", a city)
--geo-id	LinkedIn geo id, when you know it
--posted	Posting age: r86400 (24h), r604800 (week), r2592000 (month)
--remote	Workplace: 1 on-site, 2 remote, 3 hybrid
--experience	Experience level, 1..6
--job-type	Type: F, P, C, T, I, V, O
--sort	R relevance, DD date
--hydrate	Follow each stub to a full job record
--save	With --hydrate, upsert each job into the store

linkedin jobs "golang engineer" --remote 2 --posted r604800 -n 50

post

linkedin post <url> [url ...]

Fetches a single public post or article, best effort: JSON-LD first (the DiscussionForumPosting or Article node) with an Open Graph backstop. Single public posts and articles generally return data; when one is walled, linkedin exits 4.

linkedin post https://www.linkedin.com/posts/example-activity-123456789

id

linkedin id <input> [input ...]

Classifies and normalizes each argument into a (kind, id) pair without fetching. Kinds are profile, company, school, job, post, and unknown. Pure local work, never blocked.

linkedin id https://www.linkedin.com/in/williamhgates

url

linkedin url <kind> <id>

Builds a canonical LinkedIn URL from a kind and an id.

linkedin url profile williamhgates

db

Subcommand	Does
db path	Print the store file path
db count	Count stored records
db query	Read stored records back out

linkedin db count

cache

Subcommand	Does
cache path	Print the cache directory path
cache info	Location, file count, and size
cache clear	Remove every cached page

linkedin cache info

Meta

Command	Does
info	Print the resolved configuration
version	Print version, commit, and build date

linkedin info

Global flags

These apply to every command. See configuration for the full list and their defaults.

Flag	Meaning
-o, --output	Output format, auto/list/table/markdown/json/jsonl/csv/tsv/url/raw (default auto: list on a TTY, jsonl piped)
--fields	Comma-separated columns to include
--no-header	Omit the header row in table/csv/tsv output
--template	Go text/template applied per record
--color	auto, always, or never
-n, --limit	Limit number of results (0 means no limit)
-q, --quiet	Suppress progress on stderr
-v, --verbose	Increase verbosity (repeatable)
--rate	Minimum spacing between requests (default 2s)
--timeout	Per-request timeout (default 30s)
--retries	Retry attempts on 429/5xx (default 3)
--cache-ttl	On-disk cache freshness window (default 24h)
--no-cache	Bypass the on-disk page cache for this run
--refresh	Force a re-fetch and overwrite the cache
--data-dir	Root dir for cache and store (env LINKEDIN_DATA_DIR)
--cookies	Netscape cookie jar to lend a session
--dry-run	Print actions, do not perform them
--profile	Named profile to load