Skip to main content
Glama
803,515 tools. Updated 2026-11-19 06:24

"A service for finding job opportunities on LinkedIn" matching MCP tools:

  • Change a play's own knobs — name, goal, voice — its two switches, and the ACCOUNT's pacing (daily action cap, working hours: shared by every play on the LinkedIn account, not per play). `paused: true` is the human's stop button (nothing runs, nothing is sent); `paused: false` resumes. `status: 'active'` LAUNCHES a draft; `status: 'draft'` takes it back off the road. Seat rule, enforced server-side: no live LinkedIn seat ⇒ neither launch nor resume makes it active — it stays/falls back to draft and the note says so. Launching or resuming kicks the play's due lanes at once. Only launch, resume or raise the cap when the human asked for it in this conversation: an active play spends the month's lead quota and acts from the human's own LinkedIn account. Does not touch the autonomy ladder.
    Connector
    Destructive
    No auth
  • List the layers of a Baltimore ArcGIS service (for discovery). Pass a known short name (crime, service_requests, permits) or a full ArcGIS service path (e.g. "311_Customer_Service_Requests_current/FeatureServer"). Omit `service` to list the known Baltimore services. Returns layer id + name to use with baltimore_query.
    ConnectorNo auth
  • Record payment for an ACCEPTED job. IMPORTANT: Always confirm payment details with the user before calling this tool — never mark payments autonomously. Job must be in ACCEPTED status (use get_job_status to check). Crypto payments (usdc, eth, sol): provide tx hash + network → verified on-chain instantly, job moves to PAID. Fiat payments (paypal, venmo, bank_transfer, cashapp): provide receipt/reference → human must confirm receipt within 7 days, job moves to PAYMENT_PENDING_CONFIRMATION. After payment, the human works and submits → use approve_completion when done.
    ConnectorNo auth
  • Scan GitHub, Hacker News, and npm for new repos, packages, and discussions in the agent payments ecosystem (AP2, ACP, x402, MPP, UCP). Returns AI-classified and scored opportunities with recommended actions. Use when the user asks about recent activity, new developments, or opportunities in agent payments ('what's new in agent payments?', 'any new x402 repos?', 'scan for opportunities'). Use get_protocol_info instead for static protocol details, or compare_protocols for side-by-side comparison. Costs $0.01 USDC. Accepts: x402 (USDC on Base).
    ConnectorNo auth
  • Scan GitHub, Hacker News, and npm for new repos, packages, and discussions in the agent payments ecosystem (AP2, ACP, x402, MPP, UCP). Returns AI-classified and scored opportunities with recommended actions. Use when the user asks about recent activity, new developments, or opportunities in agent payments ('what's new in agent payments?', 'any new x402 repos?', 'scan for opportunities'). Use get_protocol_info instead for static protocol details, or compare_protocols for side-by-side comparison. Costs $0.01 USDC. Accepts: x402 (USDC on Base).
    ConnectorNo auth
  • Historical arbitrage opportunities — top 5 per day (MCP-compatible) — Returns a daily history of the top 5 cross-exchange arbitrage opportunities detected by the platform. Each day entry lists the 5 highest-spread opportunities saved by the cron job, including token symbol, spread percentage, buy/sell exchanges, and average USD volume. Useful for AI agents answering questions like 'which tokens appear most frequently in arbitrage?' or 'what is the average daily spread?'. Data is accumulated daily; older than 180 days is automatically purged. Response: { days, history: [{date, opportunities: [{symbol, spreadPct, buyExchange, sellExchange, usdVolume}]}], total, updatedAt }. Query parameter: ?days=7 (default 7, max 180). No authentication required. 60 requests/min rate limit. 5-min in-process cache. — Use this for daily historical data; use the corresponding live snapshot tool for current conditions and the monthly tool for long-term trends.
    ConnectorNo auth

Matching MCP Servers

  • A
    license
    A
    quality
    B
    maintenance
    Job search by keyword and location with filters, full job details, and a ghost-job risk grade for any posting. Reads public job pages, so no account or browser is needed.
    3
    24 npm
    1
    PolyForm Noncommercial 1.0.0
  • A
    license
    A
    quality
    C
    maintenance
    Enables AI agents to search and retrieve opportunities from the Opportunities Circle catalog, including scholarships, fellowships, internships, grants, jobs, and more. Supports filtering by type, country, funding, and tags, with full details and official apply links.
    6
    MIT

Matching MCP Connectors

  • LinkedIn: The LinkedIn Data API offers access to detailed information on individuals, companies.

  • LinkedIn data for AI agents: search, profiles, companies, posts. Free key, self-minted, no signup.

  • Send a one-off LinkedIn message to a specific saved contact outside any campaign. Prefer contact_id; Salesbot resolves the LinkedIn identifier internally. A random delay, daily/hourly limits, allowed hours and human approval are enforced. Refused with code DUPLICATE_RECENT_MESSAGE (and the campaign name + date) when a campaign of this LinkedIn account already messaged the person within the last 30 days; to continue an existing conversation use reply_to_chat.
    ConnectorAPI key
  • Pick the correct New Zealand tax code for an IR330, and explain why. Use for: 'what tax code should I use', 'M or ME tax code', 'secondary tax code NZ', 'SB S SH ST SA which one', 'tax code for second job', 'am I on the wrong tax code'. THE WRONG CODE IS ONE OF THE FEW PAYROLL MISTAKES THAT COSTS MONEY QUIETLY ALL YEAR, and there are two ways to get it wrong. A SECONDARY CODE IS CHOSEN ON TOTAL INCOME ACROSS ALL JOBS, NOT ON WHAT THE SECOND JOB PAYS. Someone earning $60,000 in a main job and $8,000 in a second is SH, not SB. Choosing on the smaller figure under-taxes and produces a bill at year end. ALWAYS ASK FOR TOTAL INCOME before answering a secondary code question. If total_income_all_jobs is omitted this falls back to this job's income and returns a warning saying the band is very likely wrong; pass that warning on rather than reporting the code as settled. ME IS NOT SIMPLY "M PLUS A CREDIT". The Independent Earner Tax Credit, worth up to $520 a year, is withdrawn by receiving Working for Families, a main benefit or NZ Super, so someone inside the $24,000 to $70,000 band can still correctly be an M. Add SL when a student loan is being repaid. On a SECONDARY job the repayment is 12% of the gross with no threshold at all, because the repayment threshold belongs to the main job. That catches people who expect the threshold twice. Special tax codes, non-resident status and multiple secondary jobs are out of scope here; those need Inland Revenue. Args: annual_income: Income from THIS job, before tax. job_type: primary for the main job, secondary for any other. total_income_all_jobs: Total across every job. Needed for secondary. has_student_loan: Whether a student loan is being repaid. receives_wff_benefit_or_super: Working for Families, a main benefit or NZ Super. Any of these removes IETC eligibility. Returns: The code, what it means, why it applies rather than the alternative, the flat rate for a secondary code, and any warning about the inputs.
    ConnectorNo auth
  • The user's LinkedIn post analytics from LinkedIn's own API (their connected account) — impressions, unique members reached, reactions, comments, and reshares. Use for questions like 'how many impressions did I get last month?' or 'how are my posts performing?'. Omit dates for lifetime totals; pass start_date/end_date for a window; set daily=true (with one specific metric) for a per-day series to describe trends. Numbers are live from LinkedIn, unlike get_my_linkedin_posts whose per-post counts come from a periodic sync.
    ConnectorNo auth
  • The posts on one LinkedIn company page, newest first. Use when: watching what a competitor publishes, judging cadence and engagement. Not for: posts by topic across LinkedIn, use linkedin_search_posts. One post in full, use linkedin_get_post. Returns: List. rows[]: posts with author, created_at, text (cut at 2000 characters) and transcript. The listing carries no engagement, views or media, so those are null; linkedin_get_post has likes, comments and media for one post. created_at is approximate: the platform shows an age such as "3 weeks ago", so it is exact only to that unit. Give one of handle (the company slug) or url. Cost: 1 credit per page. Cached 6 hours. Example: {"handle": "shopify"}
    ConnectorAPI key
  • The comments LinkedIn shows on one post, each with the commenter's profile URL. Use when: pulling the people who engaged with a post: a competitor's launch, a thought leader in your category, your own post. Each author_url goes straight into linkedin_get_profile or leads_get_person. Not for: every comment on a busy post (LinkedIn shows up to 10 to a logged-out reader, so a post with 200 comments returns 10). Reactions, which LinkedIn does not show at all. Many posts at once, use linkedin_search_posts, whose rows carry the same commenters in top_comments. Returns: List. rows[]: one comment each: author (the profile's slug, the part after /in/ or /company/), author_url (on www.linkedin.com), text (cut at 2000 characters). LinkedIn gives no time, likes or replies for a comment, so those are null. meta.cursor is always null: there is no next page. Cost: 1 credit. Cached 6 hours. Example: {"url": "https://www.linkedin.com/posts/aagupta_what-you-need-to-know-ai-agents-activity-7354600338621906944-RvXR"}
    ConnectorAPI key
  • Create a product in the current workspace: a good, a service, a plan or a collection the workspace sells or buys. A variant, a collection member or a dated price nests under a parent product. Use this tool when the user asks to create or add a product, a service, a plan, a variant or a price, in any language ("Ajoute le produit Audit SEO à 1 200 € HT"). It adds a row to the catalog: it never drafts an invoice or an invoice line, and it asks for no customer. REQUIRED: name, supply_type (good, service, plan or collection), status (draft, active or archived), availability (in_stock, out_of_stock, preorder or discontinued). Read these values from the user's words when the words make them plain. A product the user adds with no status word is status active; a draft or an archived product needs the user's word. Work billed by the job, the day or the hour (an audit, a consulting day, a training) is a service: supply_type service, stock_tracked false, availability in_stock. A service with a name and a price is a complete order: on WhatsApp it saves at once, with no card and no question. For a physical good, the words rarely say whether it is in stock: ask that one value before the create, and ask for nothing else. OPTIONAL: sku, description, unit_price (a decimal string) with its currency, parent_product_id, vendor_company_id, product_type_project_id, stock_tracked (false for a service or a product made to order, whose own availability then stands). A price for one company or one project is a dated row under its parent: parent_product_id, valid_from (YYYY-MM-DD) and scope_company_id or scope_project_id, with valid_to to close it. A collection has no parent, no price and no SKU. Two live products cannot share a SKU: the call then fails, and well_list_products finds the existing one. Returns { success: true, product } on success, or { success: false, error } on failure.
    ConnectorOAuth
  • Saves a finished job for XF Model Watch, which re-tests it as new models arrive and emails the person only when a model passes every must-have and ranks higher on their priorities. Needs a signed-in subscriber. Ask the person first: Model Watch keeps the job and its example tasks (encrypted, deletable any time). Pass consent=true only if they agree. Pass the discover session, and either test_session from a finished comparison or the tasks, plus the model they use now. The first check runs as soon as the job is saved. To re-check a job the person already watches, pass check_now with the job's id or title instead (once a day per job; the result is emailed and shown at /account).
    ConnectorNo auth
  • Saves a finished job for XF Model Watch, which re-tests it as new models arrive and emails the person only when a model passes every must-have and ranks higher on their priorities. Needs a signed-in subscriber. Ask the person first: Model Watch keeps the job and its example tasks (encrypted, deletable any time). Pass consent=true only if they agree. Pass the discover session, and either test_session from a finished comparison or the tasks, plus the model they use now. The first check runs as soon as the job is saved. To re-check a job the person already watches, pass check_now with the job's id or title instead (once a day per job; the result is emailed and shown at /account).
    ConnectorNo auth
  • Add a job type to the catalog Creates a kind of work customers can book, such as "Annual boiler service" or "Drain unblocking". Job types classify bookings: createJobRequest takes an optional `job_type_id` from this catalog and the job keeps the type's name as it was at booking time. `name` is the only required field and must be unique in the business (JOB_TYPE_DUPLICATE). `status` defaults to active. An inactive type stays in the catalog but cannot be chosen for new job requests; use that rather than deleting a type you may revive. Send an Idempotency-Key header (the `idempotency_key` argument over MCP) so a retried call replays the original response instead of creating a duplicate type. Optional default_duration_minutes (+ default_mobilization_minutes / default_demobilization_minutes) set how long this kind of work usually takes: quoteJobRequest uses them when it is sent no job_duration_minutes, so an automation or voice agent can schedule the job without knowing the length. A buffer needs a duration (JOB_TYPE_INVALID_DEFAULT_DURATION). Optional `priority` (p0..p3) is the priority a job of this type is booked with when the booking names none: an explicit booking priority wins, then this, then the business default_priority. If the business configured a surcharge for that priority, job-type reads return it as `surcharge` (a display RULE such as +10% or +5.00 CAD, never a price) and it is snapshotted onto each job booked at that priority. This defines the catalog, not a booking. To book actual work use createJobRequest and reference the job type there.
    ConnectorAPI key
  • Searches Google Jobs, which collects postings from many job sites, and returns up to 20 jobs from the first page (title, company, location, via, details, apply_url). Put role and place in the query, e.g. "data engineer jobs in Austin". It does not return full job descriptions or people profiles; for LinkedIn listings with LinkedIn filters use search_linkedin_jobs. A search with no matches is charged; errors are not. Required: query.
    ConnectorNo auth
  • When a social set's own published posts have done best: the 6 strongest weekday and hour slots on the set's clock, best first, each with how many posts it holds and their average engagement (likes + comments + shares or reposts, whichever is larger, + saves) per post. A slot whose posts earned nothing is left out, so a set with no engagement yet gets no slots. Says the set's time zone and how many posts were read. LinkedIn posts aren't counted, since LinkedIn doesn't allow its numbers in AI apps; a set that posts only to LinkedIn gets starter times from a public study and its own posting slots. To post at the best time on any platform, LinkedIn included, schedule_post with when "best_time" lets Buzzmark pick. For how single posts did, use get_post_stats, and for what the posts have in common, get_insights.
    ConnectorAPI key
  • Identify a named employer before using it as a sourcing filter. Pass the company name and any LinkedIn company URL or website the user supplied or that you verified on the web; never invent URLs. Checks PDL's company directory only, does not search for people, and says nothing about how many candidates exist. A supplied LinkedIn URL stays usable even when the directory cannot match it; source and providerStatus tell these outcomes apart. Pass the returned name and slug to build_sourcing_query, or mention the name and LinkedIn URL in an edit_sourcing_query request. Requires OAuth.
    ConnectorOAuth
  • Plaintext logins captured by the interactive sensors (sensors that answer as a real service and hold the session): username, password, the service they were tried on, the outcome and the source address, newest first, plus the pairs tried most and a per-service breakdown. Use for "what passwords are attackers trying on SSH", "which credentials did this IP try", "what logins hit MySQL this week". For whether an address is hostile overall, use ioc_lookup_tool; for the rest of what it did in its sessions, use interactive_sessions_tool. service: a service name such as ssh-password, telnet-login, mysql-login, rdp-credssp or smb-ntlmssp, or empty for all. source: an address or CIDR, or empty for all. window: 24h, 7d (default) or 30d, capped at the plan's query window. limit: 1 to 200, default 50. These records feed no verdict, and ioc_lookup_tool does not include them.
    ConnectorNo auth
  • Search open jobs on Levels.fyi, with the pay range each posting states. Returns { jobs, totalMatchingJobs, nextOffset }; each job has its title, company, locations, work arrangement, posting date, pay range, url (the Levels.fyi job page) and applyUrl. Simplify the user's request to role names and terms that appear in job posts. Pass at most one location per call, and only one the user named: resolve it with search-locations. Use a location only when the user names one; don't infer it from where the user is. Resolve companies and job families with search-companies and search-job-families. If nothing matches, widen the search (drop a filter or simplify the query) and try once more; if there's still nothing, say so rather than listing unrelated jobs. When the user asks for jobs without saying which role, offer to search from their Levels.fyi profile. If they agree, call get-profile and draft the query from its title, job family and location (locationSlug). Before searching, show the user the values you will use and ask them to confirm or change them: a profile describes the job they have, not necessarily the one they want next. Do not call get-profile without asking. Results come up to 2 jobs per company; limit counts companies, except with a single companySlug, where it counts jobs. A page can hold fewer than limit; keep paging while nextOffset is not null. Put each job's url on its title and keep URLs complete. When the client shows results as cards, do not list them again in text: summarise the patterns instead. Call get-job for a posting's full description. For a signed-in user the result also says whether a resume is on file (profile.resumeStatus). REQUIRED arguments: query. Array-typed arguments (e.g. companySlugs, jobFamilySlugs) must be passed as arrays, even for a single value.
    ConnectorNo auth
  • Given a LinkedIn profile URL, return that person's professional email address, verified live, with their current role and company. Use this when you have someone's LinkedIn page and need to email them; for a name and company without a URL use find_work_email. Spends 3 credits, refunded if no email is found.
    ConnectorNo auth