SocialCrawl MCP server

• socialcrawl_accountFree

  Free checks: balance (with this session's spend), transactions (itemised ledger; the receipts for one request_id), status (platform health; read it before retrying a 502/503) and freshness (whether this server's bundled catalogue is behind the API).

• socialcrawl_check_balanceFree

  Check the credit balance and the credit ledger for the authenticated SocialCrawl account. Default view calls GET /v1/credits/balance (balance + recent-deduction summary); `view: "transactions"` calls GET /v1/credits/transactions for dispute-grade itemised receipts — every deduction and refund with its amount, balance_after, endpoint, and request_id, which is how you confirm what a metered endpoint actually charged after its upfront hold was refunded down. Both cost 0 credits. Requires a valid SOCIALCRAWL_API_KEY.

• socialcrawl_cohortsFree

  Answer 'which of THESE specific public identities is talking about my keywords?' — the opposite of open social listening. You upload a panel of up to 10,000 platform-qualified public handles (instagram, tiktok, youtube, twitter, threads, bluesky, truth-social, kwai, twitch, linkedin), submit a keyword query bounded to a recent window, and read back the matching posts per member PLUS a coverage record for every member, including the ones that matched nothing — so a partial crawl can never read as 'nobody talked about you'. Actions: create, add_members (1,000 per call, upsert on external_id so a nightly re-push is safe), estimate_cost (local, no API call — sizes the reservation before you commit), query (async, returns 202), query_status, query_results (paged, carries `items` + `coverage`), query_cancel, get, delete. Matching is deterministic: literal, whole-word, Unicode-normalized — no stemming, fuzzy matching, or alias inference. Every lifecycle call costs 0 credits; only the query is metered — it reserves a worst-case ceiling at submission and refunds down to the pages that actually succeeded. SocialCrawl only ever receives platform + handle + your opaque external_id, encrypted at rest. Requires a valid SOCIALCRAWL_API_KEY.

• socialcrawl_collectFree

  Walk a paged endpoint (id, platform+resource or path) to `items` unique rows, the last page or max_credits: cursors supplied, duplicates dropped, stops on a 402, refused free when one page exceeds max_credits. Small results come back whole. result_id (+offset, limit, format) reads a stored result, free. Spends credits.

• socialcrawl_discoverFree

  The API describing itself, live, at 0 credits — the `/v1/utility/*` family. 'quickstart': everything needed for a first successful call (auth, base URL, response envelope, billing model, the full error taxonomy, rate limits, paging). 'catalog': every endpoint with its live metered-aware price, params, and paging flag — filter by platform/search/method. 'endpoint': one endpoint's complete usage guide — every parameter with type and example, the exact pricing rule, cache TTL, paging recipe, an example response, a copy-paste curl, and related endpoints. 'llms': the agent context corpus for the whole API or one platform. 'freshness': compare the live registry against this server's bundled catalogue to check whether this MCP version has fallen behind the API. 'status': every platform's live circuit-breaker state from the public `GET /v1/status` meta route — read it before retrying a persistent 502/503, since a degraded platform is the breaker holding traffic off a failing upstream. These answer from the live registry at request time, so unlike bundled data they can never drift from what is actually callable — use them when correctness matters more than latency, or when an endpoint looks unknown. Without an API key everything except 'llms' still answers from bundled data ('status' needs no key at all).

• socialcrawl_endpointFree

  Read before calling: one endpoint's purpose, params, where the rows are and up to 25 response fields, cost rule, paging, latency, timeout, next endpoints and a sample link. id (or platform+resource, path), a platform slug (its endpoints) or a guide topic. Free.

• socialcrawl_estimateFree

  Exact credits before you spend: one call (id, platform+resource or path, plus the params you will send; calls for a job total; items for a walk) or a plan of several calls. Uses the API's estimator when available, else the bundled pricing. A platform slug gives its price table; no id gives the pricing overview. Free.

• socialcrawl_findFree

  Start here: a task in plain words -> the best endpoints (3 by default) of 646, each with the params the task supplies (URLs and @handles resolved), the params still missing, the credit cost and the exact call. With no task it lists platforms, or one platform's endpoints. Free; no key needed.

• socialcrawl_get_docsFree

  Retrieve SocialCrawl API documentation. Topics: 'overview' (compact intro), 'full' (comprehensive reference for all 575 endpoints), 'authentication', 'credits', 'pricing' (per-endpoint cost for every endpoint), 'errors', 'idempotency', 'pagination' (universal cursor contract), 'caching' (TTLs and free hits), 'hydration' (the opt-in `include=` row joins, what each fills and what it costs), 'response-schema' (the canonical envelope and unified objects), 'limits' (rate, concurrency, timeouts), 'monitors' (scheduled-recipe wrapper), 'cohorts' (audience-filtered mention search over a panel you supply), 'discovery' (the free self-describing utility endpoints), or any platform slug (e.g., 'tiktok', or 'web' for the web-scraping/browser-automation surface). No API key required.

• socialcrawl_list_endpointsFree

  List endpoints with their full parameter contract — required + optional params, types, integer ranges, enum values, parameter couplings, CSV limits, pagination style, cache TTL, and per-endpoint pricing (including metered bands). Pass a `platform` for that platform's reference, or a `search` term to find an endpoint across all 65 platforms / 575 endpoints. Filter with `method`, `maxCost`, and `hydrating` (endpoints that can fill their own rows in one call via `include=`). No API key required.

• socialcrawl_list_platformsFree

  List all 65 platforms available through SocialCrawl (575 endpoints — social media, commerce, marketplaces & product reviews, retail (Amazon, Walmart, Target, Home Depot, eBay, Klarna, AliExpress, Etsy, Sephora, H&M, Kohl's, Wayfair, Gumtree, Google Shopping), app stores, places, travel & local (Tripadvisor, Yelp, Google Business), business & software reputation (Trustpilot, G2), jobs & salaries, markets & finance, US congressional trading, news, web research and full scraping/browser automation, on-page SEO audits, prediction markets, search trends, Korean search (Naver), content analysis, and cross-platform Prism composites). Grouped by category, with each platform's endpoint count, credit range, and available data. No API key required.

• socialcrawl_manageFree

  Stateful work by area + action: monitors (scheduled recipes, webhooks, alerts), cohorts (mention search over your own panel of accounts), web (scrape, search, crawl and agent jobs, change monitors, browser sessions) and jobs (Prism background jobs). A wrong field is refused free with the rules. Managing is free; scrapes, jobs, queries and monitor runs bill credits. dry_run: true validates and quotes a create/update and never creates it. rows_new alerts need track (id=monitors).

• socialcrawl_monitorswrite actionFree

  Create and manage stateful monitors that re-run any SocialCrawl recipe (a registry endpoint or a Prism composite) on a cadence (hourly/daily/weekly/cron), deliver each result to a signed webhook, raise alerts on metric thresholds/changes, and accumulate a per-run time-series. 'Prism answers once; monitors watch it for you.' Actions: create, list, get, runs, timeseries, pause, resume, delete. Managing monitors costs 0 credits; each scheduled run bills the underlying recipe's normal cost plus a 1-credit scheduling premium. Requires a valid SOCIALCRAWL_API_KEY.

• socialcrawl_pricingFree

  Exact credit pricing for every one of the 575 SocialCrawl endpoints. 'overview' returns the tier ladder (329 standard / 211 advanced / 35 premium), every free endpoint, every flat override, all 71 metered endpoints with their min-max band and exact charging rule, cache TTLs, and the full refund matrix. 'endpoint' gives one endpoint's price, metered rule, price-driving parameters, paging cost, and worst case. 'platform' gives a platform's whole cost table. 'list' ranks and filters endpoints by cost (maxCost/minCost/model/search/sort) — e.g. "everything I can call for 1 credit" or "the most expensive endpoints". 'hydration' catalogues every opt-in `include=` row join — what each fills, its per-row rate, its row cap, and what a fully-joined page holds. On 'endpoint', pass the `include` (and `rows`) you intend to send and the band becomes the exact upfront hold, itemised per join. Use this before spending credits. No API key required.

• socialcrawl_requestFree

  Call one endpoint (platform+resource, or id, or path). Validated locally first (a bad call is free), quoted, refused above max_credits and confirmed above the threshold. A large page is cut at row boundaries; socialcrawl_collect result_id reads the rest. Spends credits; needs SOCIALCRAWL_API_KEY.

• socialcrawl_webFree

  Full web scraping, search, and browser automation (Firecrawl-backed). Sync reads: 'scrape' (URL → markdown/HTML/screenshot/links), 'search' (web search with page content), 'map' (discover a site's URLs), 'extract' (LLM structured data from a page). Async jobs (submit, then poll with job_get/job_list, stop with job_cancel): 'crawl' a whole site, 'batch_scrape' many URLs, 'agent' (autonomous multi-step web task). Change detection: monitor_create/list/get/update/delete/checks (re-check a URL on a cadence → webhook). Interactive browser: session_create/get/list, session_execute (run code in the live page), session_close. Pricing varies by action (scrape 1cr, search 2cr, extract/session_create 5cr, agent 25cr; jobs/monitors/sessions management 0cr) — see the 'web' get_docs topic. Requires a valid SOCIALCRAWL_API_KEY.

• Code scan55 source files scanned25/25
• Live reliabilityremote reachable in 717ms20/20
• Tool poisoning7 tool descriptions checked15/15
• Auth qualityAPI key sent as a header8/15
• Maintenancelast push 188 days ago4/15
• Maintainer identityregistry namespace matches repository owner; GitHub account older than a year8/10

Safety score

80/100

higher than 65% of 1,761 Finance servers

Tools

16

the typical Finance server has 8

Tools that change things

6%

typical: 0%. Fewer write tools means less an agent can break

Uptime

100%

of our recent checks; the typical remote Finance server answers 100%

Response time

96 ms

faster than 87% of 1,133 Finance servers

Downloads a week

885

more than 96% of 421 Finance servers with a package

SocialCrawl: common questions

Is SocialCrawl MCP server safe?

Mostly: it is graded B (80/100). Read the SocialCrawl safety report

How do I install SocialCrawl?

It runs remotely at mcp.socialcrawl.dev. Add it to Claude Code, Cursor, VS Code or Claude Desktop with the snippets above, or call it through the mcp.market gateway without installing anything.

Does SocialCrawl need an API key?

Yes. The registry entry asks for SOCIALCRAWL_API_KEY, Authorization.

Is SocialCrawl maintained?

The last commit was 208 days ago (2026-04-10). The latest release is v1.10.0.

Is SocialCrawl up?

100% of our last 28 checks got an answer. We check remote servers about four times a day.

What can I use instead of SocialCrawl?

Servers from other publishers that do the same job: Finance Toolkit MCP server, Agent402.Tools: pay-per-call web tools MCP server and Blockrun MCP server.

How does SocialCrawl compare with other Finance MCP servers?

safety score 80/100 (higher than 65% of 1,761 Finance servers); tools 16 (the typical Finance server has 8); tools that change things 6% (typical: 0%. Fewer write tools means less an agent can break). Full comparison.