Compass MCP vs Firecrawl MCP Server
A side-by-side look at the Compass MCP and Firecrawl MCP Server MCP servers: transport, auth, tools, quality signals, and the config to install either one in Claude, Cursor, Windsurf or VS Code.
Verdict: choose Compass MCP for real estate work as a local process; choose Firecrawl MCP Server for search & data extraction as a local process. You can also run both side by side in one mcpServers block.
Which MCP server should you choose?
Choose Compass MCP when
- You need dedicated capabilities in the Real Estate domain.
- You prefer a local stdio subprocess transport.
- You want tools like
compass_search_properties, compass_get_property, compass_get_saved_homes.
Choose Firecrawl MCP Server when
- You need dedicated capabilities in the Search & Data Extraction domain.
- You prefer a local stdio subprocess transport.
- Its auth model fits you: API key required.
- You have the keys it needs:
FIRECRAWL_API_KEY, HTTP_STREAMABLE_SERVER. - You want tools like
firecrawl_scrape, firecrawl_map, firecrawl_search.
Feature & specification comparison
Rows where the two servers differ are marked.
| Spec | ||
|---|---|---|
| Category | Real Estate | Search & Data Extraction |
| Quality signal | 61/100 · Good | 90/100 · Excellent |
| Install command | Confirmed | Confirmed |
| Transport | Local subprocess (stdio) | Local subprocess (stdio) |
| Auth | Not declared | API key required |
| Pricing | Not declared | Freemium |
| Required env vars | None listed | FIRECRAWL_API_KEYHTTP_STREAMABLE_SERVER |
| Verified owner | Community listing | Verified |
| GitHub stars | 0 | 7,669 |
| Directory activity | 1 views 0 installs 0 upvotes | 14 views 0 installs 0 upvotes |
| Tools | 18 | 25 |
| Full listing | View Compass MCP | View Firecrawl MCP Server |
Tools & capabilities
Compass MCP tools 18
compass_search_propertiesSearch Compass listings by location (city, ZIP, neighborhood) and optional filters. Resolves free-text via slugification into Compass's URL routing, then fetches the SSR search-results page and extracts the embedded listings array. Compass server-renders ~41 listings into that page (its `num`), and `total_items` reports the full market count. PAGINATION (issue #87): Compass no longer paginates the SSR search via any URL — `/page-N/`, `?page=N`, and `?start=N` all canonicalize back to page 1 and return the identical listings, so only the first SSR page (~41 listings) is reachable through this primitive. `offset` is honored WITHIN that page, and `next_offset` is emitted only when more listings remain within it — it is never a false cursor that re-fetches page 1. TO REACH BEYOND THE FIRST PAGE, narrow with `price_min` / `price_max` / `beds_min`/`beds_max` to bucket the result set into <~41-listing bands (price-banding), then search each band. Returns each matching listing's address, price, beds/baths, sqft, lat/lng, the Compass homedetails URL (`_lid/` form, content-addressed by `listing_id_sha`), and the stable `_pid/` URL via `property_url` and the surfaced `pid` field. The per-listing `primary_photo_url` / `primary_thumbnail_url` are omitted in the default `compact` view (they are Compass CDN URLs a model cannot see); pass `view: "full"` to get them, or `compass_get_property_photos` for the whole gallery. USE `pid`/`_pid/` FOR LONG-LIVED REFERENCES (trackers, sheets, bookmarks) — sha-based `_lid/` URLs change when a property is delisted and relisted. Use the sha-based URL to fetch the current listing record. Read-only; safe to call repeatedly.compass_get_propertyFetch a property's full Compass record. Pass either `url` (a full Compass homedetails URL or path from a compass_search_properties result) or `listing_id_sha` alone — when only the sha is supplied, the tool fetches /listing/<sha>/view, which redirects to the canonical /homedetails/<slug>/<sha>_lid/ page. Returns address, neighborhood, beds/baths, sqft, lot size (`lot_size_sqft` plus the derived `lot_size_acres` = round(sqft / 43560, 2), null — never 0 — for condos / missing lots), price + price-per-sqft, monthly charges, MLS status, amenities, schools, parcel number, and the canonical Compass URL. Also returns `extracted_features` (lake_front, hot_tub, basement, furnished, dock, community) keyword-parsed from the description. DESCRIPTION HANDLING: The raw `description` (Compass marketing copy) is omitted by default — pass `include_description: true` to keep it. `extracted_features` is always populated and usually sufficient. URL FORMS: Compass exposes two URL shapes for a listing. `_lid/` (content-addressed by `listing_id_sha`) — what this tool fetches and what `url` returns — is the form to use for reading the current listing record. `_pid/` (opaque short ID, in `property_url` and the surfaced `pid` field) is **stable across re-listings** and is the right choice for any long-lived reference (trackers, sheets, bookmarks); sha-based URLs go stale when a property is delisted and relisted. Read-only; safe to call repeatedly.compass_get_saved_homesNot yet supported — Compass renders /overview/favorites via an auth-scoped GraphQL we have not yet identified. Throws a clear error explaining the limitation.compass_get_saved_searchesNot yet supported — Compass renders saved searches via an auth-scoped GraphQL we have not yet identified. Throws a clear error explaining the limitation.compass_calculate_mortgageLocal-only mortgage payment calculator. Returns a full PITI breakdown (principal + interest, property tax, insurance, HOA, PMI) and total interest over the life of the loan. No network call — fully deterministic, safe to use for scenario comparison without burning a fetch. Provide either `down_payment` OR `down_payment_percent`; defaults to 20%. Property tax can be given as `property_tax_annual` or `property_tax_rate` (% of home price). PMI applies automatically when LTV > 80% and `pmi_rate` is provided.compass_get_price_historyFull listing history for a Compass property — Listed / Sold / Pending / Price Change / Delisted events with date, price, and MLS attribution. Returns three arrays: `events` covers this listing's events, `history` aggregates events from prior listings of the same property, and `events_normalized` merges both into a shared cross-MCP schema (`{date, type, price?, price_change_pct?, source_mls?}` with a fixed type enum: Listed | PriceChange | Pending | Contingent | Sold | Withdrawn | Relisted | Delisted). Pass either `url` (the full Compass homedetails URL or path) or `listing_id_sha` alone — sha-only calls fetch /listing/<sha>/view, which redirects to the homedetails page. Note: most of this data is already returned inline on `compass_get_property` (the events[] / history[] arrays live on the same listing record); call this tool only when you want the merged + normalized timeline. Read-only; safe to call repeatedly.compass_compare_propertiesFetch 2 or more Compass properties and align their facts side-by-side. Each target may supply `url` (a full Compass homedetails URL or path) or `listing_id_sha` alone — sha-only targets fetch /listing/<sha>/view, which redirects to the homedetails page. Returns the full per-property record per row (with `extracted_features` populated). Per-target errors are captured per-row — one bad target will not fail the whole call; a failed row carries `status` = `error_kind` (`timeout` / `bridge_down` / `protocol` / `other`), `retryable` and `error`, and the envelope reports `count` / `ok` / `errored`. Calls are concurrent, and the whole call is bounded by an overall deadline: any row still unsettled when it is reached comes back as `{ status: "pending", retryable: true, error }` with a top-level `pending` count — re-run just those targets. The raw `description` is omitted from each row by default — pass `include_description: true` to keep it. The redundant `summary` table is also opt-in via `include_summary: true` — by default only `results[]` is returned, which already carries every fact.compass_calculate_affordabilitySolve for the maximum home price you can afford under the standard 28/36 DTI rule. Inputs: monthly income, recurring monthly debts (car loans, student loans, etc.), down payment, interest rate, and optional property-tax rate / insurance / HOA / loan term. Output: max home price, the binding constraint (front-end vs back-end), and the full PITI breakdown at that price. Same math across the realty MCP cohort (@chrischall/realty-core). No network — pure local math.compass_get_property_photosThe full photo gallery for a Compass listing — every image in listing.media[]. Each entry returns the original CDN URL plus a thumbnail URL and pixel dimensions. Pass either `url` (the full Compass homedetails URL or path) or `listing_id_sha` alone — sha-only calls fetch /listing/<sha>/view, which redirects to the homedetails page. By default only photos (category 0) are returned; set `include_all_categories: true` to also include floorplans and other media. Returns `{ listing_id_sha, count, photos }`. Read-only; safe to call repeatedly.compass_healthcheckRound-trips a small public www.compass.com URL (/robots.txt) through ContextMint Bridge (your signed-in browser tab) and returns diagnostics: the bridge's role (host/peer/null), port, version, the extension link (linked / pair pending / not attached / never answered), the elapsed round-trip time, and a plain-English hint distinguishing 'bridge never came up' from 'extension not connected' from 'this browser can't serve a capability' from 'real www.compass.com-side problem'. Read-only, no auth required. Call this when a real tool fails and you want to know which hop broke.compass_get_by_addressResolve a free-text street address to a Compass listing's canonical URL and identifiers in one call. Walks three rungs: first the structured typeahead `POST /api/v3/omnisuggest/autocomplete` (the primary rung — Compass's address-suggest API, which routes around the AWS WAF that 403s the SSR free-text path, issues #78/#79), then `/homes-for-sale/?q=<address>` (the free-text rung) and — when those return no verified match — a slug-based search at `/homes-for-sale/<city-state-or-zip>/` (the search-fallback rung, issue #71). Each candidate is verified against the query (case + street-type abbreviation normalization, then whole-token equality, issue #45) before being accepted. Returns `{ url, listing_id_sha, pid, address, resolved, matched_via }` where `matched_via` is `"typeahead"`, `"freetext"`, or `"search_fallback"` so callers can see which rung found the match. When no rung matches, returns `{ resolved: false, error: "no listing matched" }` rather than leaking a wrong URL. When the lookup was blocked by a sign-in / AWS WAF challenge instead, it returns `{ resolved: false, status: "auth_required", error, hint }` — NOT a miss: sign in to compass.com in the browser and retry. The `url` is the stable `_pid/` form when Compass provides a `navigationPageLink` (preferred for trackers/bookmarks — sha URLs go stale on relisting), falling back to the `_lid/` form otherwise. Read-only; safe to call repeatedly.compass_bulk_getFetch up to 200 Compass listings in a single call. Returns one structured row per input target (no side-by-side summary table — use `compass_compare_properties` for that). Each row is either `{ listing_id_sha, url, status: "ok", property }` on success or `{ listing_id_sha, url, status, error_kind, retryable, error }` on failure (`status` = `error_kind`) — one bad target never fails the whole call. `retryable: true` (a bridge `timeout` after one retry, an unreachable `bridge_down`, issue #73) is NOT a missing listing, so retry it (a cold bridge usually succeeds on the second call) rather than concluding Compass has no record; `protocol` / `other` are real misses. Targets accept the same `url` / `listing_id_sha` shape as `compass_get_property`. Calls fan out concurrently. The whole call is bounded by an overall deadline: a slow or hung row never wedges it — any row still unsettled when the deadline is reached comes back as `{ status: "pending", retryable: true, error }` alongside a top-level `pending` count, so re-run just those targets. The envelope also reports `count` / `ok` / `errored`. `extracted_features` is populated per row. The raw `description` is omitted by default — pass `include_description: true` to keep it.- +6 more on the listing page
Firecrawl MCP Server tools 25
firecrawl_scrapeRetrieve and extract content from one supplied URL through Firecrawl. Use this when the request identifies a page and needs its content or defined fields. It can return markdown, HTML, links, screenshots, branding data, a targeted answer, or JSON matching a supplied schema; JSON is useful when the requested result has defined fields, while markdown preserves readable page content. This tool operates on a known page. For a set of pages use `firecrawl_crawl`, and to discover page URLs use `firecrawl_map` or `firecrawl_search`. Options include JavaScript render delay, cache age, main-content filtering, PII redaction, and lockdown cache-only retrieval. Browser actions may change the live page when interactive actions are enabled. Firecrawl may reuse recently indexed content instead of refetching the page, and the reuse window varies by domain. Set `maxAge: 0` to force a live fetch, or a smaller `maxAge` to bound how stale reused content may be. A successful response does not by itself confirm that the state it describes is still current. Returns the selected content formats and page metadata.firecrawl_mapEnumerate URLs indexed under one website through Firecrawl without fetching each page's content. Use this when the request asks for a site's URL inventory, when several relevant pages must be located, or when the desired page URL is unknown. An optional `search` term narrows the URL list, while sitemap, subdomain, query-parameter, and result-limit options control coverage. Returns matching URLs rather than page bodies. Retrieve one page with `firecrawl_scrape`; collect content across multiple pages with `firecrawl_crawl`.firecrawl_searchSearch web, news, or image sources and return ranked results. Operators include quoted phrases, `-term`, `site:host`, `inurl:term`, `intitle:term`, and `related:host`; the set is non-exhaustive. `includeDomains` and `excludeDomains` are mutually exclusive hostname filters; categories limit results to GitHub, research, PDF, or developer sources. For a programming question, add `categories: ["developer"]`. It searches an index of GitHub issues, merged pull requests, repository READMEs, and curated documentation sites, and returns the hits in `data.developer` beside the web results. `categories: ["research"]` restricts these web results to research-affiliated websites and returns page snippets. The `firecrawl_research_*` tools are a separate surface that searches paper abstracts and full text across biomedical (PubMed, bioRxiv, medRxiv) and arXiv literature. `scrapeOptions` can attach extracted page content; pages fetched this way use a fixed reuse window and ignore `maxAge`, so use `firecrawl_scrape` when a live fetch is required. Returns source-type result groups and usage metadata. Authenticated responses can include an `id` for optional search feedback.firecrawl_crawlStart a multi-page crawl at a website URL, poll it to a terminal state, and return the final status and collected data. Scope can be bounded with include/exclude paths, depth, page limit, subdomain/external-link controls, sitemap handling, delay, and scrape options. Crawl results can be large; use conservative limits when full-site coverage is unnecessary. Webhooks and interactive scrape actions are unavailable in safe mode. Returns the crawl ID, status, and page data.firecrawl_check_crawl_statusRetrieve the current status, progress, and available results for an existing crawl ID. This only reads Firecrawl job state and does not start or modify the crawl.firecrawl_extractDeprecated compatibility entry point. Use firecrawl_scrape once per known URL with formats: ["json"] and jsonOptions containing the prompt and schema. Use firecrawl_search or firecrawl_agent before Scrape when URLs are not known.firecrawl_agentStart an asynchronous web research job from a prompt, optional seed URLs, and an optional JSON schema. Use this for a requested synthesis across multiple sources when the task can wait for asynchronous completion. The agent can search, navigate, read pages, and assemble a structured result. This call returns only a job ID, not the research result. Read the job with `firecrawl_agent_status` until it reaches `completed` or `failed`; research commonly takes several minutes. If the job cannot finish within the task's available time, `firecrawl_search` and `firecrawl_scrape` can gather evidence synchronously.firecrawl_agent_statusRetrieve progress or final results for a `firecrawl_agent` job ID. A `processing` response is non-terminal and does not contain the final research result. Check again after 15–30 seconds until the status is `completed` or `failed`; complex jobs can take several minutes. If the job cannot finish within the task's available time, use `firecrawl_search` and `firecrawl_scrape` to complete the requested output. Returns job status, progress information, and result data when completed.firecrawl_interactOpen or reuse a live browser session to navigate a page, click controls, fill fields, or run browser code. Provide either `url` or `scrapeId`, and either a natural-language `prompt` or executable `code`; code can run as Bash, Python, or Node with a bounded timeout. This acts on the live site, so actions such as form submission can create persistent external side effects. Returns execution output, stdout/stderr, exit status, and session viewing URLs.firecrawl_interact_stopStop the live interact session associated with a `scrapeId` and release its resources. Returns a success confirmation.firecrawl_parseParse one supported document into markdown, HTML, links, summary, targeted answers, or JSON matching a schema. Supported inputs include common HTML, PDF, Word, RTF, OpenDocument, and spreadsheet files; PDF parsing can be bounded with `pdfOptions.maxPages`. Local MCP reads `filePath` from the server filesystem. Hosted MCP uses two calls: first provide `filePath` to receive upload instructions, upload locally, then call again with the returned `uploadRef`; do not send both fields together. Remote web URLs belong in `firecrawl_scrape`. Set `redactPII` to request redaction of personally identifiable information in the returned content. `zeroDataRetention` requires an eligible authenticated account; omit it for anonymous keyless use. Returns upload instructions for hosted phase one or parsed document content for the final call.firecrawl_monitor_createCreate a recurring scrape, crawl, or search monitor that compares each check with its retained predecessor. The simple form accepts `page`/`pages` or `queries` plus a plain-language `goal`; the advanced `body` form controls targets, schedule, change-tracking formats, judging, retention, webhook, and notifications. In the simple form, a `goal` is required. If `queries` contains one or more non-empty values and is supplied with `page`/`pages`, `queries` create the search target and page targets are ignored. A monitor schedules future network checks and can send configured email or webhook notifications. Returns the created monitor.- +13 more on the listing page
Ready-to-paste client configs
Paste either block (or both) into your client config, e.g. claude_desktop_config.json or ~/.cursor/mcp.json. We only show a config when the install command is confirmed.
Compass MCP
{
"mcpServers": {
"compass-mcp": {
"command": "npx",
"args": [
"-y",
"compass-mcp"
]
}
}
}
Firecrawl MCP Server
{
"mcpServers": {
"firecrawl-firecrawl-mcp-server": {
"command": "npx",
"args": [
"-y",
"firecrawl-mcp"
],
"env": {
"FIRECRAWL_API_KEY": "YOUR_FIRECRAWL_API_KEY_HERE",
"HTTP_STREAMABLE_SERVER": "YOUR_HTTP_STREAMABLE_SERVER_HERE"
}
}
}
}
Frequently asked questions
Compass MCP is categorized under Real Estate and uses a local stdio subprocess. Firecrawl MCP Server belongs to Search & Data Extraction and uses a local stdio subprocess. Pick Compass MCP for real estate work and Firecrawl MCP Server for search & data extraction.
Copy the mcpServers block from the config section on this page into your client's config (for example claude_desktop_config.json or ~/.cursor/mcp.json), then fully restart the client.
Compass MCP has no declared pricing — check the project's own site. Firecrawl MCP Server is listed as freemium — a free tier with paid plans. Auth: api key required. If a server needs environment variables such as API keys, define them under the env key in your client's MCP config.
Yes. MCP clients support multiple servers: add compass-mcp and firecrawl-firecrawl-mcp-server as separate keys in the same mcpServers object, and your assistant will call tools from either one.
On AllMCPs, Compass MCP has a quality signal of 61/100 (Good) with 1 views, 0 installs and 0 GitHub stars. Firecrawl MCP Server has 90/100 (Excellent) with 14 views, 0 installs and 7,669 GitHub stars.