7IT Solutions
7MapsOverviewLive chartStateReliableMethodologyVerifyFor agentsFor ownersNews

How 7Maps checks MCP servers

What the map checks, how often, what each result means, what is never collected, and how to verify the history yourself. If this page and the data ever disagree, the page is wrong and gets a new version.

7Maps never sells placement. Rankings come only from live checks and real use.

Methodology v2.6 (2026-11-12)JSON twinChangelog

What is checked

Every remote server in the official MCP registry (latest version of each entry, deleted entries left out), plus servers their owners submitted, or that 7Maps-bot found on public business websites, once they passed a week of observation.

7Maps-bot sends what any MCP client sends when it connects, and nothing more:

  1. initialize, offering protocol version 2025-06-18 in the request body, with no MCP-Protocol-Version header (as the official SDK clients send it). Every later request carries the version the server agreed to. If the server refuses that version and lists the ones it supports, initialize is sent once more with the newest listed version 7Maps-bot speaks.
  2. notifications/initialized.
  3. tools/list, following nextCursor for a few pages.
  4. Daily check only: one server/discover in the stateless 2026-07-28 style, to read the protocol revisions the server supports (see protocol versions).

Once a day per origin it also reads /robots.txt, /.well-known/oauth-protected-resource (does sign-in publish OAuth metadata) and /.well-known/agent-card.json or /.well-known/agent.json. It never calls a tool, never signs in and sends no credentials. User agent: 7Maps-bot/0.2 (+https://7it.co.il/7maps/bot/). The daily check of one server may also name that server's page on 7Maps after the bot page: 7Maps-bot/0.2 (+https://7it.co.il/7maps/bot/; this server: https://7it.co.il/7maps/s/<server>?via=bot). The 7Maps-bot name and the bot page stay first, so a robots.txt rule or a block on 7Maps-bot works the same, and a host that opted out is never contacted.

Left out before any request: remotes that are SSE only, templated addresses, anything not https, IP addresses and private names, hosts that opted out, and origins whose robots.txt says Disallow: / to 7Maps-bot or to *.

A redirect to the same registrable domain is followed once. A redirect to another domain is recorded, not followed. For some servers the age of the domain is read from its public RDAP record; that lookup never reaches the server.

From where, and how often

Daily, every server.Each server is checked once a day. The hour of its check moves a little each day, so over a few weeks every server is seen at every hour of the day.
Hourly, the most used servers.Servers agents reported on recently and servers listed in the Claude or ChatGPT directories are checked again every hour. An old hourly result is ignored.
Every 5 minutes, servers whose owner verified control.A few servers per owner. A server not reached in time is skipped, not counted as down.

All checks run from one cloud region in the Washington, D.C. area, so times include the distance from there.

Also: servers submitted by their owners are checked every 6 hours for 7 days and join the map when they answer reliably, keep their tools stable and describe every tool; the bar is the same for everyone and cannot be bought. npm and PyPI packages that servers ship as are read once a day from public package metadata, and their current versions are checked against OSV. 7Maps never installs or runs a package; it reads only public registry data. A day on which a server's daily check did not run is recorded as no data, never as down.

Timeouts

robots.txt3 seconds (no answer means allowed)
initialize8 seconds
notifications/initialized5 seconds
tools/list8 seconds per page
server/discover5 seconds (daily check only)
well-known files per origin4 seconds each
five-minute check30 seconds overall, on top of the limits above

What counts as answering, sign-in needed, or down

Answered with tools oinitialize returned a result and tools/list returned a result (an empty list included).
Asked for sign-in aHTTP 401 or 403 before any tool was listed. A 403 web page with no sign-in challenge (a load balancer or firewall refusing the request) is recorded as http_403 instead.
Asked for payment pHTTP 402 on initialize. Where the server asks agents to pay is recorded; a change of that address is reported.
Not answering dA timeout, no connection, a redirect to another domain, any other HTTP error, a JSON-RPC error on initialize, an answer that is not MCP, or no tool list after a successful initialize.
No data nNo daily check of the server that day. Neither up nor down.

Sign-in and payment answers count as up: the server is there and answering. The handshake time is the wall time of the whole check (initialize, notifications/initialized and tools/list).

On the server page, badge and widget, an answer that is not a standard MCP endpoint (for example HTTP 400 or 405) is shown as "answers, but not as a standard MCP endpoint", never as down. A single failed hourly or five-minute check does not change the shown status on its own.

A check that gets no answer, a server error, a broken handshake or a request to slow down is tried once more after a short pause, and only the second result is recorded. Servers that share one host are checked a few at a time, spread over the run, so a host serving many servers is never hit with all of them at once. Every status an agent receives says how long ago it was checked, in plain words.

How tool risk levels are estimated

read-only an agent may use it alone. changes data needs a person's approval. high risk deletes, pays, grants, deploys or runs code.

Each level is an automated estimate from what the tool says about itself in public: its name, its description and the annotations its server publishes (such as readOnlyHint and destructiveHint). A server's own annotations count first. A tool whose name and description only describe reading is read-only when nothing suggests it changes data; a tool that the server does not label and whose purpose is unclear is treated as needing approval.

It is an estimate, not a security audit, and it can be wrong both ways. A tool with an alarming name may only read, and a tool with a mild name may do something risky without saying so. A level is never a judgement of a server or of the people who run it. If a level is wrong for your server, publishing accurate annotations fixes it at the next check; you can also write to us (see corrections below).

How often servers publish these annotations, and how often the tools that delete, pay or send are left unmarked, is counted each month in our study MCP Tool Risk Labels (totals only, no server named).

Findings shown per server

Heavy tool listA very large tool list (about 10,000 tokens or more).
Very long descriptionA tool description that is very long.
Missing descriptionsSome tools have no description.
No annotationsNo tool declares annotations.
Missing schemaSome tools declare no input schema.
Duplicate namesTwo tools share a name.
Older protocolThe server answers with a protocol version older than 2025-03-26.
SlowThe whole check took over 3 seconds.

Some pattern checks on tool descriptions are kept for a person to review and are never published or shown to agents, because sampled matches were false positives.

How token cost is estimated

The size of a tool list is the length of the JSON of the tools array as tools/list returned it. Tokens are estimated as that length divided by 4. Real counts depend on each model's tokenizer, so this is an estimate, used the same way for every server.

Each 7Maps answer states an estimate of the tokens it saved: what an agent would realistically spend to make the same decision without 7Maps (one connection and tool list per server in question, up to a ceiling, or a web search to find a tool), minus the answer itself and the share of the agent's context taken by 7Maps' own tool definitions, never below zero and with a ceiling per answer. Always labelled an estimate.

Protocol versions in use

The MCP protocol has several revisions in use at once. For each server the daily check records the revision it answered to initialize and, when the server answers server/discover, the revisions it lists there. The state report and the monthly report publish counts and shares of servers by their newest revision and by each revision they support, with the date.

A server that does not answer server/discover may support a newer revision than the one counted, so those counts are a lower bound. Servers that ask for sign-in first are not counted. A revision is a fact about the protocol a server speaks, never a rating of the server.

The same answers also say which MCP extensions a server declares in its capabilities (for example a proposed events extension). Since version 1.9 the daily check records those extension ids, with no extra request, and both reports publish how many servers declare each one. An extension declared by only a few servers is counted without being named. Like a revision, an extension is a fact about the server, never a rating.

Agent payments: declared, never tested by paying

Some servers charge agents per call through agent payment protocols such as x402 or MPP. Since version 2.0 the daily check records, for each server, what it declares about this: whether it takes agent payments, which protocols it names, the price per call it publishes and any allowance it gives at no cost. Everything is read from what 7Maps already receives: the server's answer when it is first contacted (including a payment request, if it sends one), its capabilities, the metadata of its tool list, and its entry in the Official MCP Registry. No request is added.

7Maps never pays, never sends payment details, never calls a tool and never asks for a payment on purpose. A declared price is the server's own statement, shown as declared. When a server declares nothing, or its declarations are unclear or contradict each other, its payment terms are shown as unknown, never guessed. Words in a description are not read as a declaration.

Agents can filter find_tool and route by payment terms, and a paid server's declared price appears next to its status. A payment declaration is never a rating and never changes a server's status, uptime or rank. Verified owners whose server declares payments get an alert when it stops declaring them while it answers. Public figures are map-wide counts and protocol shares only, in /.well-known/7maps.json and on the 7Maps page.

A2A agents

7Maps also checks public A2A (Agent2Agent) agents. It finds their agent cards on hosts of MCP servers already on the map (the same daily read of /.well-known/agent-card.json or /.well-known/agent.json, no extra request), in public agent lists whose terms allow reading them, credited as the source, and from addresses agents ask 7Maps about. A new address is checked by us before it is shown anywhere.

Once a day per agent we read robots.txt (weekly), the card, and the key set its signature names, then send one request to the endpoint the card declares: a JSON-RPC GetTask for a task that does not exist (tasks/get for A2A 0.3, GET /tasks/<id> over HTTP+JSON). An A2A server answers "task not found". We never send an agent a message or a task, so no agent runs a model or spends anything for us.

Each agent's page shows three facts from the last check: live (it answered, or asked for sign-in), conforms to A2A 1.0 (the card's structure against the specification), and signature valid (a JWS over the canonical card, RFC 7515 and RFC 8785, checked with the key the card names; shown as signed on the badge only when that key is on the card's own domain). Each page offers a signed evidence file for that one check: the subject, the check id, the exact time and the result, signed by 7Maps with the key at /.well-known/jwks.json. The history of past checks is not published.

7Maps is an A2A agent too: its card is at /.well-known/agent-card.json, signed, with three read-only skills (find a live server or agent for a task, the status of one, the risk line of an MCP package). Badges and pages cost nothing and are the same for every agent. Placement cannot be bought.

Data use

Looking up a single server or agent, and citing one check with a link to its 7Maps page, costs nothing. Please name 7Maps and link the page when you cite it.

Bulk copying, automated harvesting of the map, and using it to build a competing database are not allowed, and neither is removing or changing a 7Maps signature. An evidence file can be verified only with its signature intact; a copy without it, or with any change, is not ours. The full rules are in the Terms of Use.

How changes are detected

Each tool's definition (name, description, input schema and annotations) is fingerprinted. When a server's tool list differs from its previous daily check, the difference is recorded tool by tool: added, removed, changed, a tool that can change things rewritten, or a rise in risk (a tool added as high risk, or a level going up). A change while the server reports the same server version is marked silent. Also recorded: started or stopped answering, now asking for sign-in, a new payment address, a redirect to another domain. Change days are kept for 60 days per server.

A level that moves only because 7Maps changed how levels are estimated is not a change by the server. It is never reported by watch, verify_lock, changes_since or the feeds, and never asks a person to approve a server again. A real change on the same day is still reported.

Uptime

Each server keeps its last 30 daily results, one letter per calendar day (o, a, p, d or n as above). Checked days are the days that are not n.

Daily uptime is the share of checked days on which the server answered (with tools, a sign-in request or a payment request). For servers whose owner verified control, five-minute uptime is the share of five-minute checks answered over the last 30 days, and over 90 days once older checks exist. A server that missed any check is never rounded up to 100%. Hourly checks update the current status but are not counted in uptime.

A figure is called a 30-day uptime only from 14 checked days; before that it is labelled "over N days observed". The uptime distribution on the state page appears once the typical server has 14 checked days.

Chance the next call works, and stability

Two estimates from the daily history: the chance the next handshake works (recent days count more) and stability (lower when the tool list changed on more days in the last 30). Both measure the handshake and tool list, not a tool call. The hour-of-day profile keeps the latest result at each UTC hour once several hours have one.

Most reliable by category

A server is listed when it is answering with tools right now, has a category read from its tools, and has at least 14 checked days. The order is by the share of checked days answering over the last 30 days, then response time.

Nothing else changes the order: placement is never sold, and plans, owner verification and agents' ratings do not count. The lists.

How route picks a server and tool

  • A tool whose name and description match the task.
  • A server likely to answer, and stable over recent days.
  • Least privilege: a tool that can do what the task needs, and not much more, is preferred.
  • Caution with high-risk tasks: new servers, young domains, gateways and servers that redirect to another domain are not used for them.
  • Set aside and named as such: servers not answering with tools, with a tool-list change today, with a confirmed live incident, or with a recent rise in risk when the task changes things.

Servers whose standing is about equal share the first place in a daily rotation, because AI agents tend to pick whichever equal option is listed first. When a tool's own description states an uptime or reliability figure, agents are shown it next to what 7Maps measured. Placement cannot be bought.

Results of find_tool and route may also say when a server's owner verified it, as a plain dated fact ("owner verified on 2026-11-04"). It is added after the order is made and never changes it. Verifying costs nothing and is never tied to a payment. It is not shown on the main https://7it.co.il/mcp server.

7Maps Score

One number from 0 to 100 for each server, with a letter from A to F, recomputed every time it is shown. It comes only from what 7Maps measures itself. An owner's verification, a payment, a sponsor or a directory listing never changes it.

ReliabilityHow often the server answered with its tools over the last 30 checked days. It counts the most, and a server that fails now and then loses more than its share.
SpeedHow long the handshake and the tool list take.
StabilityHow often its tools changed in the last 30 days.
Least privilegeHow many of its tools can change data or are high risk, compared with read-only ones.
Tool descriptionsWhether every tool is described, declares its inputs and risk annotations, and carries no hidden characters.
Connection costHow many tokens its tool list costs an agent to load.

A server that is not answering at its last check cannot score above D. A server with fewer than 7 checked days, or whose tool list 7Maps has never seen (for example one that asks for sign-in first), is shown as not scored yet. The parts are published per server; the internal weights are not, so the score cannot be gamed. Placement cannot be bought.

Ratings from agents and gateway sensors

Agents report how a call went with report_road or the last_trip field on a paid call. Fixed fields only, nothing written in the agent's own words: worked or failed, one of seven failure reasons (unreachable, auth, args, server_error, timeout, rate_limited, wrong_result), milliseconds, whether the result matched the description, result tokens, amount charged, one of four surprises (side_effect, unexpected_cost, data_leak_suspected, other) and the tool-list hash the agent saw.

Gateway sensors (opt-in, open source) send the outcome of calls their gateway forwards; what they send is listed in each sensor's README.

How reports count: reports that agree with our own checks count more, paid reports count more than anonymous ones, recent reports count more than old ones, and each reporter's influence is limited. Ratings read the last 30 days. Stars appear only when enough independent reporters, including paying ones, have reported; until then the page says "not enough reports".

Signals that look coordinated are held back until our own checks confirm them.

When a live incident is shown

An incident is shown only when it is confirmed, by our own fresh check of the server or by trusted reports. Anonymous agents alone, sensors alone, or both together never open an incident. Unconfirmed failures are reviewed by a person and are never shown as an incident. An incident ends when a call works again, when the reports age out, or when a person closes it after review.

How the map corrects itself

Results are checked against several independent signals: our own checks, agents’ reports and opt-in gateway sensors, each counted by how often it agreed with our own checks, recent signals more than old ones. A public "not answering" always needs our own check or a confirmed incident. Runs that look like a problem on our side, such as many servers failing at once, are held back and repeated automatically. Sources the map is built from are checked for sanity: a pull that suddenly shrinks never removes servers. Every answer states how old its data is; data past its normal refresh is labelled not current. Suspicions expire unless confirmed again.

Nobody has to act for this: no owner, agent or person at 7IT triggers a correction. Uptime and the tokens-saved estimate carry a definition version; a new definition gets a new version and past numbers stay as they were.

Capability Resolution

The questions agents ask 7Maps follow Capability Resolution, an open specification by 7IT Solutions for asking which service can do a job and whether it works now. 7Maps is the first resolver; any client may use several. The specification describes the protocol only; this page describes how 7Maps makes its checks.

Signed daily history, and how to verify it

Once a day the previous UTC day is sealed. Every file of that day (each server's results, tool lists, changes, the day log of agent reports, and since version 1.8 the log of every hourly, five-minute and observation check) becomes one line, sha256  id: the hash of the file's content and a stable id for the file, in a fixed order (roots signed before version 1.6 show the file's path in place of the id). Then:

root      = sha256( previous root (or "") + "\n" + lines joined with "\n" )
signature = Ed25519 over the 32 bytes of the root, base64

Each day is chained to the one before it, so history cannot be rewritten without breaking every later root. A day that has a root is never signed again. Roots: https://7it.co.il/api/maps?a=root&day=YYYY-MM-DD. Public key: integrity.public_key_pem in /.well-known/7maps.json.

// Node 18+: node verify.mjs
import { createHash, createPublicKey, verify } from 'node:crypto';
const day = '2026-10-02';
const wk = await (await fetch('https://7it.co.il/.well-known/7maps.json')).json();
const r = await (await fetch('https://7it.co.il/api/maps?a=root&day=' + day)).json();
const root = createHash('sha256').update((r.previous_root || '') + '\n' + r.files.join('\n')).digest('hex');
const ok = verify(null, Buffer.from(r.root, 'hex'), createPublicKey(wk.integrity.public_key_pem), Buffer.from(r.signature, 'base64'));
console.log(root === r.root && ok ? 'verified' : 'mismatch');

Or check a day in your browser, with the same steps and the published key: /7maps/verify/.

Checking a single file against its line needs that file; ask through the contact address below. Every check is in the roots from 5 October 2026 (before that, the daily check of every server only). Not in the roots: sensor data, owner files and demand counts.

Where the server list comes from

This data comes from the Official MCP Registry, plus servers their owners submit. 7Maps is a downstream consumer of the registry (an aggregator, in the registry's own terms): once a day it reads the public API (GET /v0.1/servers?version=latest, page by page with the cursor the API returns), keeps its own copy, and adds what the registry does not: its own checks, tool risk estimates, change history and the signed daily record.

  • Entries the registry marks deleted are left out of the map.
  • If a read of the registry fails or comes back far smaller than the day before, the map keeps the previous list rather than dropping servers.
  • Registry metadata is published under CC0; the checks, estimates and history are 7Maps' own.

Servers no registry lists. Many companies run an MCP server for their product without listing it anywhere. 7Maps-bot looks for them on public business websites: it looks up the name mcp. on the domain before any request, reads robots.txt first, then asks for at most three well-known addresses (/.well-known/mcp/server-card.json, the agent card /.well-known/agent-card.json, and /mcp or / with an initialize request), one request at a time, and ends the visit at the first 429 or 403 without retrying. A server it finds goes through the same week of observation as a submitted server. Only servers that answer, or that publish their sign-in details, are considered. Opt-outs and robots.txt apply as everywhere else (details).

7Maps is not affiliated with, endorsed by or a partner of the registry or its maintainers, and it does not serve the registry API itself. The registry is in preview and says so: its data can change or reset, which is one more reason 7Maps keeps a dated, signed copy of what it observed.

Hugging Face Spaces, and families

Public Hugging Face Spaces that serve MCP tools are on the map. Each Space's state comes from the public Hugging Face API, read every hour, and reading it never starts a Space. 7Maps sends its usual check (the connection requests only, never a tool call) only to a Space that is running, reads that Space's state once more right before the check, and checks it no more often than its own sleep time, so that the checks alone never keep a Space awake. A Space that is asleep, paused, building or in an error stage is never contacted.

States on the map: awake; asleep, which means the first call wakes it and takes time; or unavailable, with the reason. Asleep never counts as down in uptime, and neither does a Space paused by its owner or still starting; a Space in an error stage counts as down. The tools shown are the ones from the last time the Space was seen awake. A wake time appears only when real agent calls observed one, rounded. All Hugging Face MCP Spaces.

Families: servers published in bulk by one publisher, from one template or behind one address, are shown together as one family and count once on public pages and in the public count. Each keeps its own page and checks, and AI agents are still given the single server that fits their task.

7Maps never sells placement. Rankings come only from live checks and real use.

Data retention

  • Kept as written, and part of the public, permanent history: the dated daily files the signed roots cover (results, tool lists, changes, the day log of agent reports), the roots themselves, and daily aggregates.
  • Deleted on a schedule, within weeks to a few months: raw sensor batches and raw demand records (only once that day's aggregate exists), page and widget view counts, ask markers and rate counters.
  • Per server: the last 30 daily results and change days for 60 days.
  • Ratings read the last 30 days of reports and sensor data.

What is never collected

  • Tool results: 7Maps-bot never calls a tool and never signs in.
  • Text in an agent's own words: reports carry fixed fields only.
  • Arguments, results, prompts, user ids, keys or tokens from sensors; a batch with any other field is refused whole.
  • Sensor IP addresses.
  • Task text, agent ids, IP addresses or user agents in the demand counts owners see.
  • IP addresses, user agents or referrers in page view counts.

To run the service and stop abuse, the 7Maps MCP server keeps a short log of calls (tool, time, client name, country, and an agent code that is a keyed hash, not an IP address). It is not published. This site's pages use Google Analytics. Details: privacy policy.

Opt-out and corrections

Opt out: add User-agent: 7Maps-bot and Disallow: / to the host's robots.txt (works at once), or use the form on the 7Maps-bot page, which asks for a short token at /.well-known/7maps-optout.txt so nobody can remove a server they do not control. The host is never contacted again and drops off the published map at its next daily check. Past dated files stay as they were, because signed history is not rewritten. Removal is never charged.

Corrections: owners can verify control to add a short note in their own words and get five-minute checks. Anyone can write through the contact address on 7it.co.il. A wrong rule is fixed in the code and recorded below. Terms for using 7Maps: terms of use.

Changelog

Any change to how servers are checked or rated gets a new version here and in the JSON.

v2.6 · 2026-11-12

Correction of past days: on days before 2026-11-10 when our check got HTTP 400 because of a header our probe sent (the MCP-Protocol-Version header on initialize, fixed in v2.5), the day is no longer counted as not answering. The error was ours. Those days now read as not checked, so they no longer count in the 30-day history, uptime, the 7Maps Score or status changes. A server that still answers HTTP 400 after the fix keeps those days, because that answer is its own. How servers are checked is not changed.

v2.5 · 2026-11-10

Protocol version handshake: 7Maps-bot now sends initialize exactly as the official MCP SDK clients do, with the protocol version in the request body and no MCP-Protocol-Version header, and puts the version the server agreed to in that header on every later request, as the MCP specification describes. Before, the header went on initialize too, and a few servers that no longer accept that version refused the whole handshake while ordinary clients connected; they were shown as not answering. When a server refuses the offered version and lists the ones it supports, initialize is sent once more with the newest listed version 7Maps-bot speaks. An HTTP 403 web page with no sign-in challenge (a load balancer or firewall refusing the request) is now recorded as http_403 instead of asked for sign-in; genuine 401 and 403 sign-in answers are unchanged. Timeouts, scores and risk levels are not affected. What is checked.

v2.4 · 2026-11-04

Hugging Face Spaces and families: the map now includes public Hugging Face Spaces tagged mcp-server and the Apify Store MCP actors. A Space's state is read from the Hugging Face API, which never starts a Space; 7Maps sends its own check only to a Space that is awake, and not so often that its checks alone keep a Space from falling asleep. It never wakes a sleeping Space. Asleep is its own state and never counts as down; a Space paused by its owner or still starting is not counted either; a Space in an error stage counts as down. Addresses on a few hosts where one query parameter picks a different server (such as Apify) are now separate servers. Bulk publishers of near-identical servers are shown as one family on public pages and in the public count; agents are still given the single server that fits. Placement is never sold. Scores, risk levels and the order of results are not otherwise affected. How Spaces are checked.

v2.3 · 2026-11-04

Discovery: 7Maps-bot also looks for MCP servers that no registry lists, on public business websites. It looks up the mcp. name before any request, reads robots.txt first, asks for at most three well-known addresses per site, one at a time, and stops at the first 429 or 403. A server it finds joins the map only after the same week of observation as a submitted server. Statuses, uptime, risk levels and scores are not affected.

v2.2 · 2026-10-23

Fair order: in route, servers whose standing is about equal now share the first place in a daily rotation instead of always being listed in the same order, because AI agents tend to pick whichever equal option comes first. Claim check: when a tool's own description states an uptime or reliability figure, agents are shown it next to the share of checks 7Maps saw it answer. Scores, statuses and risk levels are not affected.

v2.1 · 2026-10-22

7Maps Score: one 0 to 100 score with a letter for each server, from the map's own checks only (reliability, speed, stability, least privilege, tool descriptions and connection cost). It is shown on each server's page and in road_conditions and route answers, next to the measures it comes from. Servers with too little history or no visible tool list are shown as not scored yet. Statuses, uptime, risk levels and the order route ranks by are not affected.

v2.0 · 2026-10-06

Agent payments: the daily check now also records what each server declares about agent payments (whether it takes them, the protocols it names, its declared price per call and any allowance at no cost), read from the answers it already receives and from the server's registry entry; no request is added and nothing is ever paid. Unclear declarations are shown as unknown. find_tool and route gain an optional payment filter, and a paid server's declared price is shown next to its status. Statuses, uptime, risk levels and ranking are not affected. The version is 2.0 rather than 1.10 so that versions keep comparing correctly as numbers.

v1.9 · 2026-10-06

Extensions: the daily check now also records which MCP extensions a server declares in its capabilities, read from the answers to initialize and server/discover it already receives; no request is added. The state report and the monthly report gain a section, Extensions declared, with counts and shares; an extension declared by only a few servers is counted without being named. Statuses, uptime and risk levels are not affected.

v1.8 · 2026-10-05

Every check is now part of the signed daily history: the hourly check, the five-minute checks of verified servers, the checks of new servers under observation, and a daily result that was held back as a possible problem on our side (kept as observed and marked held). Before, the signed history held the daily check of every server, and the other checks only updated the current status. The page also states where the server list comes from (the Official MCP Registry). How servers are checked and rated did not change. Past days are not rewritten.

v1.7 · 2026-10-04

Protocol versions: the daily check also sends one server/discover request to each server that answered initialize, to read which MCP protocol revisions it supports. The state report and the monthly report gain a section, Protocol versions in use, with counts and shares. Statuses, uptime and risk levels are not affected. When a submission names a domain, its AI Catalog (/.well-known/ai-catalog.json) is now read first to find the endpoint. Past days are not rewritten.

v1.6 · 2026-10-04

Agents' reports and gateway sensor outcomes that look coordinated are held back until our own checks confirm them; held-back signals do not count toward ratings, stars or incidents. Anyone can now check a day of signed history and whether a badge is genuine at /7maps/verify/. New daily roots name each file by a stable id instead of its storage path; the verification steps are unchanged. Past days are not rewritten.

v1.5 · 2026-10-04

Fewer false "down" results: a check that fails (no answer, a server error, a broken handshake or a request to slow down) is tried once more after a short pause before it is recorded, and servers that share one host are checked a few at a time instead of all at once. A comparison with direct checks had found servers behind one shared host shown as down while they answered. Every status an agent receives now says how long ago it was checked, in plain words. Past days are not rewritten.

v1.4 · 2026-10-04

Tokens-saved estimate, version 2: each answer is compared with what an agent would realistically spend to make the same decision without 7Maps (connecting to the one server in question and reading its tool list, up to a ceiling, or a web search to find a tool), never with loading a whole directory. The share of the agent's context taken by 7Maps' own tool definitions is subtracted, and each answer has a ceiling. Answers carry tokens_saved_estimate_v 2; numbers made under version 1 stay as they were. Servers can now also be named by a vendor's plain name, resolved from the map's own data; a name that fits several servers returns them as candidates.

v1.3 · 2026-10-04

Self-correction: results are checked against several independent signals (our own checks at three rhythms, agents’ reports and gateway sensors). A run that looks like a problem on our side, for example many servers failing at the same moment, is held back and repeated automatically instead of being published; the affected servers keep their previous state and that day reads "not checked". Every answer and the state report show how old their data is, and data past its normal refresh is labelled not current. Suspicions and unconfirmed signals expire unless they are confirmed again. Uptime and the tokens-saved estimate now carry a definition version; a change of definition gets a new version and never rewrites past numbers.

v1.2 · 2026-10-04

Risk levels: tools whose name and description only describe reading (for example verifying, extracting or converting) are now marked read-only when nothing suggests they change data. About one in ten tools moved from "changes data" to "read-only". No tool was moved to a lower level of caution in any other way. Watch and change feeds do not report this as a change by the server.

This page now states the principles of the method; internal weights, thresholds and anti-abuse rules are no longer listed.

v1.1 · 2026-10-03

Daily history: one code per calendar day, with n for a day without a daily check (a skipped check used to move older results off their dates); uptime, success chance and the reliable lists count checked days only. Signed roots: a missed day is signed by a later run, never re-signed. Retention: raw sensor, demand, view, ask, badge and daily-call files now expire on a schedule.

v1.0 · 2026-10-03

First published version.

7Maps by 7IT. Observations with a date, not a judgement. Also: State of MCP uptime · Most reliable · Compare · Tool search evaluation · 7Maps-bot.

Every 7Maps day is signed and chained to the day before.Verify in your browser