overassessed MCP server
ai.agentlookups/overassessedagentlookups.aistreamable-httpv1.0.0Scanned
Maryland property-assessment fairness checks vs similar homes, with appeal windows and links.
1 tool: 1 change things.
- Reads private data · 0
- Reads text others write · 0
- Can send data out · 0
- 1 finding
Tools listed without sign-in. Remote posture
With Anlyon
1 tool here changes things. Govern it.
- Anlyon lists the tools and adopts none.
- Each tool you adopt gets approvals and receipts.
- Your agent never holds the key.
A tool your agent calls directly is not routed through Anlyon and is not governed by it. Anlyon Radar only lists and labels, and these labels are the same either way.
What the overassessed MCP server can do
The overassessed MCP server is published by agentlookups.ai, a domain the MCP Registry verified. It has 1 tool. It can create or change data (1 tool: check_assessment), but cannot delete data. None of its tools only reads.
None of its tools reads private data, brings in text other people write, or sends data out.
The fixed checks matched 1 finding (long-description). Each one comes with the text it matched.
Its remote endpoint lists its tools without sign-in, and its tools changed once since Anlyon Radar first saw them.
Last scanned Nov 17, 2026, version 1.0.0.
Tools
Grouped by what each tool does, most consequential first. Open one for its reasons and full description; descriptions are the publisher's own words.
1 tool
Changes things 1Claude Code: ask
check_assessmentFairness check for one Maryland property assessment against genuinely similar homes (same triennial reassessment group, similar size, lot and age, same zip), from a copy of the sta…1 finding
Why it is in this group
- Nothing in its name or description says what it does, so it is counted as a write.
- The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
Findings
- Info
The description is 5,987 characters long, over the 2,000 this check looks for.
What this check looks forMatched text
description ends: "...ot an appraisal, and not legal advice; the official SDAT record is authoritative."
Description, as the server wrote it Fairness check for one Maryland property assessment against genuinely similar homes (same triennial reassessment group, similar size, lot and age, same zip), from a copy of the state's own assessment roll (SDAT) refreshed monthly. Input is a Maryland street address with its zip, its Maryland city or county, or MD, or an SDAT account id (acct); the city, county, state and zip can be written in the address itself (1 STATE CIR, ANNAPOLIS, MD 21401), a city or county narrows the match to that place the way a zip does, and MD alone matches the street across the whole state. A single match returns the parcel record, a uniformity object, a plain-language verdict, Maryland's appeal windows with official filing links, and honesty notes. In the parcel record, market_value is the full value an appeal argues about; taxable_now is the phased-in assessment for the Maryland tax year holding today (tax years run July 1 to June 30; taxable_now_tax_year gives its from and to dates, YYYY-MM-DD), and taxable_next with taxable_next_tax_year is the one for the tax year after; either pair is left out when the roll does not hold that year. The uniformity object gives BandSize (how many similar homes), where the assessment sits among them as percentiles (PctFMV for market value, PctImprPSF for improvement value per square foot, PctLand for land value), their medians (MedianFMV, MedianImprPSF, MedianLand), the median sales ratio of assessment to sale price (MedianSalesRatio) and its coefficient of dispersion (COD) over RatioSales sales dated RatioFromYear through RatioToYear (three calendar years ending with the year of the roll's newest sale dated on or before today), and Branch. RatioSales counts sales of $100,000 or more that the roll codes as arm's-length sales of one improved parcel; foreclosures, gifts, auctions, sales of several parcels for one price and sales of a vacant lot are left out. When the roll records no lot size or no year built for the home, the band skips that test: NotCompared names it (lot size or year built; both as lot size, year built) and the plain verdict's not_compared field says so in words. NotCompared is absent when every test ran. When the assessment sits high (PctFMV 75 or more) or low (below 25) and the home's rank on total value per square foot sits 3 or more in 10 from its rank on total value, the plain verdict's size field says so: total value is the rate per square foot times living area, so the difference comes from the home's size. Verdict branches: fair means the assessment is below the 75th percentile of similar homes (the plain headline calls it on the low side below the 25th and lower than almost all of them below the 10th, and the normal range otherwise); review means it is on the high side; elevated means it is higher than almost all of them; insufficient means fewer than 25 similar homes exist for a fair comparison, which is no finding either way. On the insufficient branch the percentile and median fields are 0, meaning not computed. A single-match or multiple_matches result also carries a roll object: as_of is the date the state last updated the roll and pulled is the date this copy was taken (both YYYY-MM-DD), overdue is true when this copy is more than 35 days old, and a field the server does not know is left out. MedianSalesRatio and COD are 0 when RatioSales is below 10, meaning no ratio is published. When several parcels share the address, the result is multiple_matches, listing parcels with their account ids; passing one of those ids as acct selects that parcel. The list is in unit order, each option carries unit (the unit as the roll writes it, such as UNIT 204) when the roll records one, and total gives how many parcels share the address; when more than 10 do (a condominium building, for instance), the list holds the first 10. A unit written after the street (100 HARBORVIEW DR UNIT 204, 100 HARBORVIEW DR #204, or as its own comma-separated part) narrows the match to that unit when the roll records units at the address, and a single match then carries parcel.unit; when the roll records units there and none is the unit written, the result is an error naming that unit, with no match list; a unit at an address where the roll records none is not applied. Every refusal and every miss below is a tool error: isError is true and the text content is the error message, with no verdict and no match list. When nothing matches, the result is an error; the roll writes addresses in its own form (e.g. 1 STATE CIR), so a miss can mean a different spelling, a town other than the one the roll records, or a zip other than the one the roll records, rather than a missing parcel. A zip is 5 digits; a ZIP+4 (21202-1234) is read as its first five digits, and a zip in the zip input takes precedence over one written in the address. A malformed zip or acct returns an error naming the expected form, a zip outside Maryland returns an error stating that only Maryland is covered, an address naming another state or territory returns an error stating that only Maryland is covered and naming that state as not covered yet, an address with no Maryland zip, city, county or state returns an error stating that only Maryland is covered and that a zip, city or county is needed, and an acct the roll does not hold returns an error stating that no parcel has that account id; none of these carries a verdict or a match list. Without a Maryland zip or MD, an address also gets the error asking for a zip, city or county when a comma-separated part after the street is not read as a Maryland place, a state, a zip, a unit or the US, such as a town Maryland lacks, another country or a state written as Miss. (100 MAIN ST, DOVER, KENT or 100 MAIN ST, LAUREL, MISS.), or when its county is written only as <name> Co with no other Maryland place (1 MAIN ST, HOWARD CO). The check is a fairness snapshot of the public record, not an appraisal, and not legal advice; the official SDAT record is authoritative.
Annotations the server declares
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
- readOnlyHint: true
- title: Check fairness of a Maryland property assessment
This is not a safety rating, and no finding is not a safety guarantee. The groups and findings come from fixed checks of the tool definitions the server returned on the date shown, not from its code or what it does when called. How the checks work.
Exfil surface
Which tools read private data, which read text someone else can write, and which can send data out. One of each is a path a prompt injection can use.
No private-data, untrusted-input, outbound path in this server alone: no tool is tagged private-data or untrusted-input or outbound.
- Reads private data0
- Reads files, mail, messages, databases, secrets, calendars or repositories.
- Reads text others write0
- Brings in text someone else can write: web pages, inbound email, messages, issues and comments.
- Can send data out0
- Can carry data to people or systems: it sends, posts or publishes, or takes a recipient, channel or URL.
Other servers in the same agent add their own tools to these paths. Audit your whole setup to see them together.
Remote posture
What anyone on the internet sees at the remote endpoint: an anonymous initialize and tools/list, and the public OAuth metadata.
Looked at Nov 20, 2026, 19:23 UTC, unchanged since Oct 22, 2026. No credentials were sent and no tool was called. What is looked at.
- Reachable
- yes
- HTTP status of an anonymous initialize
- 200
- tools/list served without credentials
- yes
- Tools that change things, listed without credentials
- 1
- Protocol version
- 2025-06-18
- WWW-Authenticate on 401
- not observed
- resource_metadata in WWW-Authenticate
- not observed
- RFC 9728 protected resource metadata
- not observed
- Authorization servers
- not observed
- Legacy HTTP+SSE transport
- no
- Dynamic client registration
- not observed
- Client ID metadata documents
- not observed
- WarnThe whole server
The endpoint served tools/list without credentials, and the list has 1 tool that changes things (1 write). Whether a call also works without credentials was not tested: tools are never called.
What this check looks forMatched text
https://overassessed.agentlookups.ai/mcp: tools/list returned 1 tool to an anonymous client, including check_assessment (write)
Findings
Matches of fixed checks, with the text they matched.
- Infocheck_assessment
The description is 5,987 characters long, over the 2,000 this check looks for.
What this check looks forMatched text
description ends: "...ot an appraisal, and not legal advice; the official SDAT record is authoritative."
History
Changes to its tool definitions and to its remote endpoint, newest first.
Tools1 changed.~1 changedView diff
- Changed:check_assessmentChanges thingsdescriptioninputSchema
Install
Pinned to version 1.0.0. Values in angle brackets are yours to fill in.
claude mcp add --transport http overassessed https://overassessed.agentlookups.ai/mcpclaude mcp add --transport http overassessed https://overassessed.agentlookups.ai/mcpRun in a terminal.
Claude Code permissions
Allow the reads, ask before a change or a way out, deny the deletes. Paste into .claude/settings.json.
{
"permissions": {
"allow": [],
"ask": [
"mcp__overassessed__check_assessment"
],
"deny": []
}
}Tools that only read are allowed, tools that change things ask first, tools that delete or cancel are denied. A read that can send data out (it takes a URL, say) asks first too. It assumes the server is installed as overassessed. Review it before you use it.
Pin
Keep the decision: fail CI when the tools change, follow every change, or show a badge.
Pin in CI
Fail a build when the tools you reviewed are not the tools the server serves today.
curl -s "https://radar.anlyon.com/api/v1/servers/ai.agentlookups/overassessed/verify?hash=af24b4cfaff8237eba2d7ad0a1bc5fae06f2f153753dcd1e6780a6628e05817a" | jq -e '.changed == false'Pinned hash af24b4cfaff8, from Nov 17, 2026, 07:17 UTC.
Watch
Get told when its tools change: a new tool, a new description, a new class.
https://radar.anlyon.com/servers/ai.agentlookups/overassessed/changes.xmlAn Atom feed. Works in any feed reader or Slack's RSS app.
Badge
[](https://radar.anlyon.com/servers/ai.agentlookups/overassessed)Source: MCP Registry. Synced Oct 21, 2026, 17:41 UTC.
overassessed MCP server: questions and answers
Answered from the scan above, not from the publisher's description.
Is the overassessed MCP server safe?
- No scan can promise that. What its definitions show: 1 tool, 1 of which can change, send or delete something, and 1 finding from fixed checks. Allow the reads, ask before each change, and deny the deletes; the permissions block on this page does that.
Can the overassessed MCP server delete data?
- No tool is classed as destructive: going by their definitions, none of them deletes, cancels or revokes anything.
Can the overassessed MCP server do things that cannot be undone?
- No tool is classed as irreversible: none of them sends, pays, publishes or merges, going by their definitions.
What tools does the overassessed MCP server have?
- 1 tool: 1 write. check_assessment.
Does the overassessed MCP server need sign-in or OAuth?
- Its remote endpoint listed its tools to an anonymous client, without sign-in.
Does the overassessed MCP server send data to outside services?
- None of its tools sends, posts or publishes, or takes a recipient, channel or URL.
Has the overassessed MCP server changed its tools?
- Yes, once since Anlyon Radar started tracking it, most recently on Nov 3, 2026. Every change is on this page and in its Atom feed.