linkedin-relay-mcp 3.24.0 → 3.26.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/.claude/skills/linkedin-relay/SKILL.md +51 -13
- package/README.md +5 -2
- package/dist/cli.js +1382 -717
- package/dist/cli.js.map +1 -1
- package/dist/index.js +18 -9
- package/dist/index.js.map +1 -1
- package/dist/mcp-shim.js +1164 -487
- package/dist/mcp-shim.js.map +1 -1
- package/package.json +1 -1
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
---
|
|
2
2
|
name: linkedin-relay
|
|
3
|
-
description: Research LinkedIn from the user's own account — find people by company, past company, school, place, title, degree or whose connections they are, search posts, companies, groups, events, courses, products, services and schools, hunt jobs with every filter LinkedIn has and collect their apply links, read profiles, read the chronological feed — and, through the lnrelay CLI, post (text, image or video), comment, reply, react, delete and edit the user's own profile (headline, About, photo, cover image, roles, skills) under the approval mode the owner chose. Use when looking at the user's own network (connections, followers, who they follow, invitations they sent, who viewed their profile) or their saved items (saved posts, courses, the job tracker), when finding or evaluating people (who works or worked somewhere, alumni, people in a role or a city, someone's connections), companies, groups, events, courses, posts on a topic, roles or job openings on LinkedIn, when the user wants job listings or application links, when reading what the user's network is posting, when the user wants to read their LinkedIn messages or draft replies to them, or when the user asks to post, engage or update their profile on LinkedIn. Cannot send messages, connect, or apply to jobs: those do not exist.
|
|
3
|
+
description: Research LinkedIn from the user's own account — find people by company, past company, school, place, title, degree or whose connections they are, search posts, companies, groups, events, courses, products, services and schools, hunt jobs with every filter LinkedIn has and collect their apply links, read profiles, read the chronological feed — and, through the lnrelay CLI, post (text, image or video), comment, reply, react, delete and edit the user's own profile (headline, About, photo, cover image, roles, skills) under the approval mode the owner chose. Use when looking at the user's own network (connections, followers, who they follow, invitations they sent, who viewed their profile, their notifications) or their saved items (saved posts, courses, the job tracker), when finding or evaluating people (who works or worked somewhere, alumni, people in a role or a city, someone's connections), companies, groups, events, courses, posts on a topic, roles or job openings on LinkedIn, when the user wants job listings or application links, when reading what the user's network is posting, when the user wants to read their LinkedIn messages or draft replies to them, or when the user asks to post, engage or update their profile on LinkedIn. Cannot send messages, connect, or apply to jobs: those do not exist.
|
|
4
4
|
---
|
|
5
5
|
|
|
6
6
|
# linkedin-relay
|
|
@@ -118,6 +118,17 @@ and `lnrelay job 4479248573 4475225930 4476083905`.
|
|
|
118
118
|
name, profile id, profile URL, headline, connection degree, `role` ("Job poster") and `verified`.
|
|
119
119
|
It costs one page call per job (60 a day), so ask only for the finalists. `[]` means LinkedIn
|
|
120
120
|
shows no one for that job, which is most jobs; `hiringTeamError` means it could not be read.
|
|
121
|
+
- **How crowded a job is, and the company behind it (Premium): `job` with `insights:true`**
|
|
122
|
+
(`--insights`). Each job gains `premium`: `applicants` {`total` who clicked apply, `lastDay`,
|
|
123
|
+
`seniority` [{level, share, count}], `education` [{degree, share}]} and `company` {`name`,
|
|
124
|
+
`headcount` {asOf, now, sixMonthsAgo, yearAgo, twoYearsAgo, `growth1y` in %}, `function` (the
|
|
125
|
+
job's own function there: name and the same counts), `medianTenure`, `competitors` [{name,
|
|
126
|
+
industry, url}], and LinkedIn's generated `hiring`, `priorities` and `landscape` text}. One more
|
|
127
|
+
job read per job (60 a day), so finalists only. `premiumError` says why it could not be read.
|
|
128
|
+
- **Jobs where the user would stand out (Premium): `jobs` with `topApplicant:true`**
|
|
129
|
+
(`--top-applicant`). LinkedIn's own list of jobs where it says the user would be a top
|
|
130
|
+
applicant: job cards as `jobs` returns them, 100 a call, usually all in one. It takes no
|
|
131
|
+
keywords, location or filters; rank and read finalists as with any search.
|
|
121
132
|
|
|
122
133
|
- **Put every keyword set in one `jobs` call**, never one call per keyword: the tool merges them by
|
|
123
134
|
job id, so you read each job once and `job` reads it once. Separate calls leave the merging to you
|
|
@@ -271,9 +282,9 @@ Every filter LinkedIn offers for this query, each with its options' `value`, `la
|
|
|
271
282
|
|
|
272
283
|
### `profile` — mid. Read only the finalists.
|
|
273
284
|
```
|
|
274
|
-
profile id:"<public-id | profile id | urn | linkedin.com/in/… URL>" [contact] [compact] [fields:"a,b"]
|
|
285
|
+
profile id:"<public-id | profile id | urn | linkedin.com/in/… URL>" [contact] [posts:N] [skills] [recommendations] [compact] [fields:"a,b"]
|
|
275
286
|
```
|
|
276
|
-
CLI: `lnrelay profile <id|url> [--contact] [--compact | --fields a,b,c]`. The whole profile page: `name`,
|
|
287
|
+
CLI: `lnrelay profile <id|url> [--contact] [--posts N] [--skills] [--recommendations] [--compact | --fields a,b,c]`. The whole profile page: `name`,
|
|
277
288
|
`headline`, `about`, `location`, `industry`, the top card (`title`, `company`, `school`), and every
|
|
278
289
|
section that has entries:
|
|
279
290
|
- `experience`: every role, newest company first: `title`, `company`, `companyId` (what `company`
|
|
@@ -288,7 +299,11 @@ section that has entries:
|
|
|
288
299
|
|
|
289
300
|
Descriptions make a full profile long. **`compact:true`** turns each role, school and entry into
|
|
290
301
|
one line and drops descriptions (About stays); `fields:"name,experience"` keeps only those.
|
|
291
|
-
|
|
302
|
+
|
|
303
|
+
**`recommendations:true`** (`--recommendations`) adds `recommendations`: `received` and `given`, each
|
|
304
|
+
`[{name, url, headline, degree, context, text}]`, where `context` is the date and how they worked
|
|
305
|
+
together, in the user's LinkedIn language. 15 a page, both ways: 2 page calls for most people. The
|
|
306
|
+
texts are long; use `fields` to keep only what you need. It failing leaves `recommendationsError`.
|
|
292
307
|
|
|
293
308
|
**`contact:true`** (`--contact`) adds `contact`, the profile's Contact info as it shows to the user,
|
|
294
309
|
for one more profile read: `email`, `websites` [{`url`, `type`}], `phones` [{`number`, `type`}],
|
|
@@ -296,6 +311,15 @@ for one more profile read: `email`, `websites` [{`url`, `type`}], `phones` [{`nu
|
|
|
296
311
|
own name. In the profiles read, only a connection showed an email. It failing leaves `contactError`
|
|
297
312
|
and the rest of the profile. Ask for it only when the user wants to reach someone.
|
|
298
313
|
|
|
314
|
+
**`posts:N`** (`--posts N`) adds `posts`, their N newest posts, as `search posts` rows: the posts
|
|
315
|
+
search filtered to posts from them, newest first, on the search budget (1 call per 20). It failing
|
|
316
|
+
leaves `postsError`. For their posts on a topic, use `search posts` with `from:[id]` and keywords.
|
|
317
|
+
|
|
318
|
+
**`skills:true`** (`--skills`) reads every skill in place of the top 20, from the profile's Skills
|
|
319
|
+
page: 1 page call per 10 skills (47 skills is 6 calls). Use it only when the skills decide
|
|
320
|
+
something; `counts.skills` says how many there are. It failing keeps the top 20 and leaves
|
|
321
|
+
`skillsError`.
|
|
322
|
+
|
|
299
323
|
### `feed` — mid. What the network is posting.
|
|
300
324
|
```
|
|
301
325
|
feed [limit:1-20]
|
|
@@ -431,18 +455,32 @@ With a list, its rows, 50 a call: people (connections, following, followers) as
|
|
|
431
455
|
does neither.
|
|
432
456
|
- Not available: received invitations, newsletters, "people you may know".
|
|
433
457
|
|
|
458
|
+
### `notifications` — 1 page call per 25. The bell.
|
|
459
|
+
```
|
|
460
|
+
notifications [unread] [limit:N] [compact]
|
|
461
|
+
```
|
|
462
|
+
CLI: `lnrelay notifications [--unread] [--limit N] [--compact]`. Newest first: `unseen` (the bell's
|
|
463
|
+
count), and items as `kind` (LinkedIn's own name, e.g. `FOLLOW_PAGE_HERO`, `LEARNING_RECS_FOR_JOB_SEEKERS_BY_STAGE_INAPP`), `text`,
|
|
464
|
+
`detail`, `at` (ISO), `read`, `url` (where it goes), `from` (who or what it is about).
|
|
465
|
+
- Reading here does not mark them seen; opening the page in the browser does.
|
|
466
|
+
- Not available: the page's tabs (my posts, mentions, jobs). Filter on `kind` instead.
|
|
467
|
+
|
|
434
468
|
### `viewers` — 2 page calls. Who viewed the user's profile.
|
|
435
469
|
```
|
|
436
470
|
viewers [compact]
|
|
437
|
-
|
|
438
|
-
|
|
439
|
-
`
|
|
440
|
-
|
|
441
|
-
|
|
442
|
-
`
|
|
443
|
-
|
|
444
|
-
|
|
445
|
-
|
|
471
|
+
viewers days:365
|
|
472
|
+
```
|
|
473
|
+
CLI: `lnrelay viewers [--days 7|14|28|90|365] [--compact]`. As the user's profile-views page shows
|
|
474
|
+
it: `total`, `privateViewers` (private-mode viewers are a count, no names), and `items`: named
|
|
475
|
+
viewers as `name`, `id` (what `profile`, `messages` and `connectionOf` take), `url`, `degree`,
|
|
476
|
+
`headline`, `viewed` ("2d ago"); anonymous ones as `anonymous: true`, `description` ("Founder in the
|
|
477
|
+
Software industry from Budapest"), `viewed`, `searchUrl`.
|
|
478
|
+
- **Premium** names every viewer in the range and takes `days` (default the page's 90; 365 is the
|
|
479
|
+
longest). It adds `highlights` [{label, value}]: top location, industry and company, and
|
|
480
|
+
`companies` [{name, share}]: where viewers work, with their share in percent.
|
|
481
|
+
- A free account is shown up to 3 viewers and the total is labelled with its `period`.
|
|
482
|
+
- `meta.state` is `complete` only when the named and private viewers add up to `total`.
|
|
483
|
+
- Degree, labels and times come in the session's language, which may not be English.
|
|
446
484
|
|
|
447
485
|
### `inbox` — 1 page call per 20 conversations. The user's messages.
|
|
448
486
|
```
|
package/README.md
CHANGED
|
@@ -39,7 +39,7 @@ claude mcp add linkedin-relay -s user -- bun "$(npm root -g)/linkedin-relay-mcp/
|
|
|
39
39
|
|---|---|
|
|
40
40
|
| `people` · `people-filters` | people search with every filter: company, past company, school, place, degree, title, someone's connections |
|
|
41
41
|
| `search` · `search-filters` | posts, companies, groups, events, courses, products, services, schools — each with its filters |
|
|
42
|
-
| `profile` · `feed` | a profile in full — name, about, every role with dates, education, skills and the other sections
|
|
42
|
+
| `profile` · `feed` | a profile in full — name, about, every role with dates, education, skills and the other sections; with `--contact` its contact info, `--posts N` their newest posts, `--skills` every skill, `--recommendations` those received and given — and your feed |
|
|
43
43
|
| `jobs` · `job` · `job-filters` | job search with every filter, several keyword sets merged in one call, and apply links |
|
|
44
44
|
| `post` · `reactions` | a post's comment thread, and who engaged |
|
|
45
45
|
| `scheduled` · `experience` · `skills` | your queued posts, and your own roles and skills |
|
|
@@ -47,7 +47,8 @@ claude mcp add linkedin-relay -s user -- bun "$(npm root -g)/linkedin-relay-mcp/
|
|
|
47
47
|
| `company` | a company page in full: about, website, industry, size, employees, followers, headquarters, specialties, locations, affiliated and showcase pages — and, with `--posts N`, its newest posts |
|
|
48
48
|
| `events` · `event` | the events page's recommended and Premium events, and any event in full: when, where, attendees, organiser, its site |
|
|
49
49
|
| `network` | your My Network: connections, following, followers, pages, events, groups, invitations sent — or the count of each |
|
|
50
|
-
| `viewers` | who viewed your profile: the total, the viewers LinkedIn names (3 on a free account), anonymous ones, the private-mode count |
|
|
50
|
+
| `viewers` | who viewed your profile: the total, the viewers LinkedIn names (3 on a free account, all with Premium), anonymous ones, the private-mode count; with Premium, `--days` up to 365 and the top location, industry and companies of your viewers |
|
|
51
|
+
| `notifications` | your notifications, newest first, read or not — without marking them seen |
|
|
51
52
|
| `inbox` · `messages` · `draft` · `drafts` | read your messages without marking them read; draft replies locally — nothing is ever sent |
|
|
52
53
|
| `sync` · `local` · `connections` · `my-posts` | fill a local cache, then query it for free |
|
|
53
54
|
| `whoami` · `doctor` · `risk` · `budget` · `cache-status` | session, standing and spend |
|
|
@@ -154,6 +155,8 @@ lnrelay jobs "frontend engineer" "AI agent developer" "agent engineer" \
|
|
|
154
155
|
--location "European Union" --workplace remote --posted week --limit 200 --compact
|
|
155
156
|
lnrelay job <id> <id> <id> # finalists in full, apply links
|
|
156
157
|
lnrelay job <id> --team # + who posted it, who is hiring
|
|
158
|
+
lnrelay job <id> --insights # Premium: who applied, company headcount and insights
|
|
159
|
+
lnrelay jobs --top-applicant --compact # Premium: jobs where you'd be a top applicant
|
|
157
160
|
```
|
|
158
161
|
|
|
159
162
|
- **Every filter the jobs page has:** location or exact geo id, distance, date posted, experience,
|