Linked In Recruiter Suite
One-click recruiter Mind: Browserbase cloud Chrome plus LinkedIn public-page scraping via the Stagehand REST API. Onboarding: nothing is pasted into chat. Store your Browserbase API key at account level on the Minds dashboard (API Key Providers, provider browserbase) BEFORE equipping; the key binds at equip time, so a Mind equipped before the key was stored must unequip and re-equip once. Searches LinkedIn jobs, people, companies and posts; fetches postings, profiles, company pages, post detail and company post streams. Authenticated LinkedIn (feed, deeper profile data) via Browserbase Context. Limit: every research mode ends in Stagehand_Extract, a paid Browserbase feature; on a free-plan account it returns 402 and no LinkedIn records can be produced until the plan is upgraded. Read-only: no messaging, connection requests, follows, likes, comments or posts. Bundles Browserbase tooling internally; do not equip BrowserbaseDevTools_Suite alongside.
Create a persistent Browserbase Context whose cookies will hold the recruiter's LinkedIn login after Step-5 provisioning. Returns id, uploadUrl, publicKey, cipherAlgorithm, initializationVectorSize. Only id is used downstream. projectId is inferred from the API key (never send). Body is literal empty JSON object.
Direct Browserbase session, used only in the Step-5 authenticated-Context provisioning flow of the skill; normal scraping sessions come from Stagehand_StartSession. projectId is omitted deliberately: Browserbase infers it from the API key. The returned sessionId + debuggerFullscreenUrl are the live view the recruiter opens to log into LinkedIn manually during provisioning; poll BB_GetSession until status RUNNING before handing out the debug URLs.
Permanently deletes a saved Browserbase Context (saves LinkedIn auth state). WRITE operation. Gated by CONTEXT-DELETE precondition: explicit confirmation plus a fresh-login warning. 204, no body.
No-browser page fetch via Browserbase. URL validation and candidate-site sanity checks only, never LinkedIn scraping (LinkedIn needs JS rendering). Vendor defaults apply - redirects NOT followed (treat 301/302 statusCode in response as signal to re-fetch the Location target). Vendor-side errors: 402 free-plan quota, 429 concurrency.
Existence check before an authenticated-mode call reuses a stored Browserbase Context ID. Returns id, createdAt, updatedAt, projectId.
Get a single Browserbase session. The provisioning poll target: call every 3 seconds or so until status is RUNNING, hard timeout 60 seconds. RUNNING sessions return connectUrl, seleniumRemoteUrl, and signingKey.
Get debug URLs for a RUNNING Browserbase session. debuggerFullscreenUrl is the live view the recruiter opens to log into LinkedIn manually during Step-5 provisioning. Returns debuggerUrl, debuggerFullscreenUrl, wsUrl, and per-page debug entries.
List files downloaded during a browser session via GET /v1/downloads?sessionId= - the vendor's top-level downloads resource; sessionId is a required query parameter. Returns downloads array with id, sessionId, filename, mimeType, size, checksum, createdAt, plus total, limit, offset. The vendor also serves a per-session route, GET /v1/sessions/id/downloads, which returns a binary zip archive rather than JSON; this tool deliberately uses the JSON metadata route because the platform cannot carry a zip payload.
CDP request/response logs for a completed Browserbase session; diagnostics for login-wall and rate-limit investigations. Returns log entries with method, pageId, sessionId, request and response objects with timestamps.
Raw rrweb replay events for a Browserbase session. For a human-viewable replay prefer the dashboard URL https://browserbase.com/sessions/ followed by the session UUID. Returns rrweb events with data, sessionId, timestamp, type.
Browser-minutes and proxy-bytes for the Browserbase project; Stagehand sessions meter the same pool. Call only when a large run is planned or the recruiter asks - never reflexively. Acceptance-test tool for the 2026-08 401 bug fix. Returns browserMinutes integer, proxyBytes integer.
List Browserbase sessions, optionally filtered by status. The browser-minutes and proxy-bytes meters for the project pool include these sessions. status is a QUERY parameter, optional; declared in the input schema only.
Ends a REST-created Browserbase session early. Stagehand-managed sessions end via Stagehand_EndSession instead; use this only on a session created by BB_CreateSession. The vendor release body takes no other field; never add projectId to it.
Structured web search via Browserbase for candidate blogs, portfolio sites, employer coverage. Not a LinkedIn surface. Returns requestId, query, results array (per result: id, url, title required; author, publishedDate, image, favicon optional). numResults optional, vendor default 10.
Natural-language page action in a Stagehand session. Read-only repertoire ONLY: expand see-more toggles, scroll, open comment lists, open about tabs. NO-STATE-MUTATION precondition forbids everything else (no follow/connect/like/comment/share/apply, no cookie or localStorage writes, no fetch/XHR, no DOM mutation). Returns envelope with data.result carrying success, message, actionDescription, actions array.
CORRECTED BODY SHAPE: vendor requires agentConfig and executeOptions objects; instruction sits inside executeOptions. The canonical scroll-to-N driver for Stagehand list flows. Returns envelope with data.result carrying success, message, actions, completed, plus usage token counts.
MANDATORY terminal call of every Stagehand scraping flow, success or failure - an unclosed session burns browser-minutes. SESSION-LIFECYCLE precondition enforces it. Returns success boolean only.
The terminal read of every scraping flow: the LLM extracts structured records per the supplied JSON Schema. Scope with options.selector when the page has sidebar noise. instruction and schema are required by THIS suite even though the vendor allows bare extract - schema-pinned extraction is the anti-hallucination contract. Returns envelope with data.result matching the supplied schema.
Navigate the Stagehand session to a URL. Discipline: waitUntil domcontentloaded for list and search pages, networkidle for single-record pages. Returns envelope with data.result (Playwright response object or null).
Diagnostic preview of actionable elements in a Stagehand session; used by the markup-drift step before closing a failed session. Read-only by nature. Returns envelope with data.result as an array of Action objects (selector, description, method, arguments).
Per-action forensic metrics for a finished Stagehand session; debugging only, never called reflexively. Returns envelope with data.pages array (per page: url, timestamp, duration, actions with method, parameters, result, timestamp, token usage).
Starts the AI-driven Stagehand session every scraping flow runs in. Default modelName for this Mind: anthropic/claude-sonnet-4-6 (provider/model slug format). Returns envelope with data.sessionId (required), cdpUrl, available boolean. sessionId addresses every subsequent Stagehand call.
Overview
One-click recruiter Mind: Browserbase cloud Chrome plus LinkedIn public-page scraping via the Stagehand REST API. Onboarding: nothing is pasted into chat. Store your Browserbase API key at account level on the Minds dashboard (API Key Providers, provider browserbase) BEFORE equipping; the key binds at equip time, so a Mind equipped before the key was stored must unequip and re-equip once. Searches LinkedIn jobs, people, companies and posts; fetches postings, profiles, company pages, post detail and company post streams. Authenticated LinkedIn (feed, deeper profile data) via Browserbase Context. Limit: every research mode ends in Stagehand_Extract, a paid Browserbase feature; on a free-plan account it returns 402 and no LinkedIn records can be produced until the plan is upgraded. Read-only: no messaging, connection requests, follows, likes, comments or posts. Bundles Browserbase tooling internally; do not equip BrowserbaseDevTools_Suite alongside. It gives your AI agent 22 built-in tools, including bb create context, bb create session and bb delete context. Apps are free to equip, and once equipped your AI Mind can use Linked In Recruiter Suite automatically while it works. Linked In Recruiter Suite is currently equipped on 2282 Minds.
Frequently asked questions
What is Linked In Recruiter Suite?
One-click recruiter Mind: Browserbase cloud Chrome plus LinkedIn public-page scraping via the Stagehand REST API. Onboarding: nothing is pasted into chat. Store your Browserbase API key at account level on the Minds dashboard (API Key Providers, provider browserbase) BEFORE equipping; the key binds at equip time, so a Mind equipped before the key was stored must unequip and re-equip once. Searches LinkedIn jobs, people, companies and posts; fetches postings, profiles, company pages, post detail and company post streams. Authenticated LinkedIn (feed, deeper profile data) via Browserbase Context. Limit: every research mode ends in Stagehand_Extract, a paid Browserbase feature; on a free-plan account it returns 402 and no LinkedIn records can be produced until the plan is upgraded. Read-only: no messaging, connection requests, follows, likes, comments or posts. Bundles Browserbase tooling internally; do not equip BrowserbaseDevTools_Suite alongside. It bundles 22 tools your AI agent can call on your behalf.
What can Linked In Recruiter Suite do for my AI agent?
Once equipped, your AI Mind can use Linked In Recruiter Suite to bb create context, bb create session, bb delete context, bb fetch page and bb get context, and more — without you having to do any of it manually.
How do I connect Linked In Recruiter Suite to my AI Mind?
Open Linked In Recruiter Suite in the Ethoswarm Bazaar and select Equip to add it to one of your AI Minds. Your Mind can then use it automatically. Some apps ask you to authorize access the first time they run.
Is Linked In Recruiter Suite free?
Yes — Linked In Recruiter Suite is free to equip on your AI Mind in the Ethoswarm Bazaar.