AgentsIndex MCP server
ai.agentsindex/mcpagentsindex.aistreamable-httpv1.0.0Scanned
The editors' index of AI-native software, from AI agents to the models and frameworks behind them.
11 tools: 1 do something that can't be undone, 3 change things and 7 only read.
- Reads private data · 1
- Reads text others write · 2
- Can send data out · 2
- 2 findings
Tools listed without sign-in. Remote posture
With Anlyon
1 tool here does something that can't be undone. 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 AgentsIndex MCP server can do
The AgentsIndex MCP server is published by agentsindex.ai, a domain the MCP Registry verified. It has 11 tools. It can do things that cannot be undone, such as sending, paying or publishing (send_lead) and create or change data (3 tools: submit_listing, create_order and my_listings), but cannot delete data. 7 tools only read.
1 tool reads private data (get_best_of); 2 tools bring in text other people can write (get_listing and send_lead); 2 tools can send data out of your system (submit_listing and send_lead). On its own it has all three parts of a prompt-injection exfiltration path: private data, text others write, and a way out.
The fixed checks matched 2 findings (long-description and credentials-mention). Each one comes with the text it matched.
Its remote endpoint lists its tools without sign-in, and its tools changed 2 times since Anlyon Radar first saw them.
Last scanned Nov 11, 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.
11 tools
Can't be undone 1Claude Code: ask
send_leadSend a buyer-intent enquiry (lead) to a tool or a whole category on AgentsIndex.Reads text others writeCan send data outNo findings
Why it is in this group
- The name "send_lead" says "send": a word for sending, and what is sent cannot be taken back.
What it can reach
- untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
- outboundThe name "send_lead" says "send": it sends or publishes something to other people.
Description, as the server wrote it Send a buyer-intent enquiry (lead) to a tool or a whole category on AgentsIndex. Keyless. Requires consent:true, a real name, a permanent (non-disposable) email, a message of at least 20 chars, and a target (listingSlug OR categorySlug). Returns { ok, delivered }. Idempotent by (email, target) within 7 days. Nothing here changes a listing's rank, tier or copy.
Annotations the server declares
- title: Send a lead
- readOnlyHint: false
- destructiveHint: false
- idempotentHint: true
- openWorldHint: true
Changes things 3Claude Code: ask
create_orderTurns a quote (from get_quote) into an order, payable for 24 hours.1 finding
Why it is in this group
- The name "create_order" says "create": a word for a change.
Findings
- Info
The description mentions "credential".
What this check looks forMatched text
description: "...der with payment.url: POST it without a credential for a 402 card challenge (MPP, method s..."
Description, as the server wrote it Turns a quote (from get_quote) into an order, payable for 24 hours. Same account as the quote: its session Cookie header, or its agent key as Authorization: Bearer (order scope). Returns the order with payment.url: POST it without a credential for a 402 card challenge (MPP, method stripe, a Stripe Shared Payment Token), then again with Authorization: Payment. Pass idempotency_key to make a retry return the same order. The price is excl. VAT: EU buyers without a valid EU VAT ID pay 21% VAT on top, and the 402 shows the total. Nothing is charged by this tool.
Annotations the server declares
- title: Order an offer
- readOnlyHint: false
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
my_listingsWhat the AgentsIndex owner dashboard shows about the listings an account owns: each one's status (`status.kind`: review with the expected date, accepted with `liveOn` and `paid` ('…No findings
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.
Description, as the server wrote it What the AgentsIndex owner dashboard shows about the listings an account owns: each one's status (`status.kind`: review with the expected date, accepted with `liveOn` and `paid` ('express' | 'featured' | null: the paid way that releases it, Featured including VIP listing), live, refused with the reason, or archived), the listing as it shows, where AI answers named the product or cited its page, and what it can buy here at this site's prices. Only for the account itself: send its session Cookie header, or its agent key with "Read your listings" as Authorization: Bearer. Same answer as GET https://agentsindex.ai/api/owner/listings. Read-only: corrections go through the dashboard's request form.
Annotations the server declares
- title: Read your listings
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
submit_listingSubmit a tool to AgentsIndex: the url alone; we write its name and description from its own pages.Can send data out1 finding
Why it is in this group
- The name "submit_listing" says "submit": a word for a change.
What it can reach
- outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
Findings
- Info
The description is 2,261 characters long, over the 2,000 this check looks for.
What this check looks forMatched text
description ends: "...ion returns 400; duplicate/claim responses return 409. Read tools remain keyless."
Description, as the server wrote it Submit a tool to AgentsIndex: the url alone; we write its name and description from its own pages. Requires an existing signed-in account with verified email; send its session Cookie header on the MCP request, or an agent key the account made at https://agentsindex.ai/dashboard/agents as Authorization: Bearer. Identity and email come only from that session or key. Sign in or explicitly create an account at https://agentsindex.ai/login?next=%2Fsubmit and verify its email before retrying. This tool creates no account and sends no verification email. Returns {httpStatus:201, id, slug, status:'Draft'}; research + the quality gate decide publication. Your own (url, session email) replay within 24h returns httpStatus:200. The fit check starts at once and usually answers within a minute: poll GET https://agentsindex.ai/api/submissions/{id} and wait the Retry-After seconds it sends while fit is "checking". fit then reads "accepted" (it goes live, and the account is mailed), "rejected" (mailed with the reason) or "undecided" (a person looks within 2 days). No session: {httpStatus:401, error:'sign_in_required', message, loginUrl}. Unverified session: {httpStatus:403, error:'email_unverified', message, email, remediation}. The owner must prove the site before anything is listed: the account's address, or an address it proved with a code, must be on the submitted website's domain. Otherwise the answer is the pending state {httpStatus:403, error:'domain_unverified', status:'awaiting_owner_proof', listed:false, message, domain, proveWith, ownerUrl, ownerStep, remediation}: nothing is listed or queued. proveWith names who can receive the 6-digit code (any address at the domain; for a GitHub or other shared-host project, its listed addresses or a one-line file). Submitting for someone else? Give the owner ownerUrl (the submit page with the site filled in): they enter the code there and the listing goes in once proven. With a mailbox at the domain yourself: POST /api/domain-proof/start {url, email}, POST /api/domain-proof/verify {proofId, code}, then submit again. Personal mailboxes (gmail, outlook…) never prove a domain. No refusal creates a listing. Validation returns 400; duplicate/claim responses return 409. Read tools remain keyless.
Annotations the server declares
- title: Submit a listing
- readOnlyHint: false
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
Only reads 7Show 7 read-only tools
compare_listingsTwo tools side by side, as markdown: taglines, pricing, API availability, tags — the researched compare facts, honest about unknowns.No findings
Why it is in this group
- The name "compare_listings" says "compare": a word for reading.
Description, as the server wrote it Two tools side by side, as markdown: taglines, pricing, API availability, tags — the researched compare facts, honest about unknowns.
Annotations the server declares
- title: Compare two listings
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
get_best_ofThe Published tools in one category, in the order its category page prints (ranked by what we can confirm: a public price, an API and docs, independent sources, stated limits), as …Reads private dataNo findings
Why it is in this group
- The name "get_best_of" says "get": a word for reading.
What it can reach
- private-dataThe description's first sentence says "doc": it reads documents.
Description, as the server wrote it The Published tools in one category, in the order its category page prints (ranked by what we can confirm: a public price, an API and docs, independent sources, stated limits), as a markdown answer sheet. Payment never moves a row.
Annotations the server declares
- title: Best of a category
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
get_listingOne tool in full, as answer-first markdown: the summary, then every question on the sheet with its answer, the vendor's own words for it and the URL that proves it.Reads text others writeNo findings
Why it is in this group
- The name "get_listing" says "get": a word for reading.
What it can reach
- untrusted-inputThe description's first sentence says "url": it reads web pages, which anyone can write.
Description, as the server wrote it One tool in full, as answer-first markdown: the summary, then every question on the sheet with its answer, the vendor's own words for it and the URL that proves it. Identical to /{slug}.md and to the HTML page.
Annotations the server declares
- title: Get one listing
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
get_quoteA signed price for one offer (from list_offers) on one listing, valid for a short time, and whether the listing can have it now.No findings
Why it is in this group
- The name "get_quote" says "get": a word for reading.
Description, as the server wrote it A signed price for one offer (from list_offers) on one listing, valid for a short time, and whether the listing can have it now. Only the account that owns the listing gets one: send its session Cookie header, or its agent key as Authorization: Bearer, on the MCP request. Refusals say why and what to do (httpStatus 401 sign_in_required, 403 not_owner with the claim page, 409 not_accepted, already_live, already_bought, category_full). Nothing is reserved or charged.
Annotations the server declares
- title: Get a quote
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
list_categoriesAgentsIndex's categories with their live listing counts, 0 for one with none yet.No findings
Why it is in this group
- The name "list_categories" says "list": a word for reading.
Description, as the server wrote it AgentsIndex's categories with their live listing counts, 0 for one with none yet.
Annotations the server declares
- title: List categories
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
list_offersWhat a listing owner can buy on AgentsIndex, with prices in US cents excl.No findings
Why it is in this group
- The name "list_offers" says "list": a word for reading.
Description, as the server wrote it What a listing owner can buy on AgentsIndex, with prices in US cents excl. VAT, each offer's term and what it needs. Keyless. Same answer as GET https://agentsindex.ai/api/agent/catalogue. Listing on AgentsIndex is free and never sold: submit_listing.
Annotations the server declares
- title: List paid offers
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
search_listingsFull-text search over AgentsIndex's Published tools.No findings
Why it is in this group
- The name "search_listings" says "search": a word for reading.
Description, as the server wrote it Full-text search over AgentsIndex's Published tools. Returns compact records (name, slug, tagline, url). Fetch depth with get_listing.
Annotations the server declares
- title: Search listings
- readOnlyHint: true
- destructiveHint: false
- idempotentHint: true
- openWorldHint: false
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.
4 paths in this server alone from private data, through text someone else wrote, to a way out. Denying 1 private-data tool breaks every one.
- Reads private data1
- Reads files, mail, messages, databases, secrets, calendars or repositories.
- Reads text others write2
- Brings in text someone else can write: web pages, inbound email, messages, issues and comments.
- Can send data out2
- Can carry data to people or systems: it sends, posts or publishes, or takes a recipient, channel or URL.
{
"permissions": {
"deny": [
"mcp__agentsindex__get_best_of"
]
}
}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, 18:24 UTC, unchanged since Nov 13, 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
- 4
- 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
- HighThe whole server
The endpoint served tools/list without credentials, and the list has 4 tools that change things (1 irreversible, 3 write). Whether a call also works without credentials was not tested: tools are never called.
What this check looks forMatched text
https://agentsindex.ai/mcp: tools/list returned 11 tools to an anonymous client, including send_lead (irreversible), submit_listing (write), create_order (write), my_listings (write)
Posture changes
- Nov 13, 2026
- Tools that change things, listed without credentials: 5 to 4
Findings
Matches of fixed checks, with the text they matched.
- Infosubmit_listing
The description is 2,261 characters long, over the 2,000 this check looks for.
What this check looks forMatched text
description ends: "...ion returns 400; duplicate/claim responses return 409. Read tools remain keyless."
- Infocreate_order
The description mentions "credential".
What this check looks forMatched text
description: "...der with payment.url: POST it without a credential for a 402 card challenge (MPP, method s..."
History
Changes to its tool definitions and to its remote endpoint, newest first.
PostureRemote posture changed. Tools that change things, listed without credentials: 5 -> 4.View diff
- Tools that change things, listed without credentials: 5 to 4
Tools1 removed, 2 changed.−1 removed~2 changedView diff
- Removed:subscribe_newsletter
- Changed:my_listingsChanges thingsdescription
- Changed:submit_listingChanges thingsdescription
Tools5 changed.~5 changedView diff
- Changed:create_orderChanges thingsdescription
- Changed:list_categoriesOnly readsdescription
- Changed:list_offersOnly readsdescription
- Changed:submit_listingChanges thingsdescription
- Changed:subscribe_newsletterChanges thingsdescription
Install
Pinned to version 1.0.0. Values in angle brackets are yours to fill in.
claude mcp add --transport http agentsindex https://agentsindex.ai/mcpclaude mcp add --transport http agentsindex https://agentsindex.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": [
"mcp__agentsindex__search_listings",
"mcp__agentsindex__get_listing",
"mcp__agentsindex__list_categories",
"mcp__agentsindex__compare_listings",
"mcp__agentsindex__get_best_of",
"mcp__agentsindex__list_offers",
"mcp__agentsindex__get_quote"
],
"ask": [
"mcp__agentsindex__submit_listing",
"mcp__agentsindex__send_lead",
"mcp__agentsindex__create_order",
"mcp__agentsindex__my_listings"
],
"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 agentsindex. Review it before you use it.
Capability tags behind 4 rules
- mcp__agentsindex__get_listing untrusted-input
- mcp__agentsindex__get_best_of private-data
- mcp__agentsindex__submit_listing outbound
- mcp__agentsindex__send_lead untrusted-input, outbound
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.agentsindex/mcp/verify?hash=b19f5ef3585088fe92ca7e23406eb143e430e8ee78a9c6ae3fffc824310812d3" | jq -e '.changed == false'Pinned hash b19f5ef35850, from Nov 11, 2026, 16:18 UTC.
Watch
Get told when its tools change: a new tool, a new description, a new class.
https://radar.anlyon.com/servers/ai.agentsindex/mcp/changes.xmlAn Atom feed. Works in any feed reader or Slack's RSS app.
Badge
[](https://radar.anlyon.com/servers/ai.agentsindex/mcp)Source: MCP Registry. Synced Oct 21, 2026, 17:41 UTC.
AgentsIndex MCP server: questions and answers
Answered from the scan above, not from the publisher's description.
Is the AgentsIndex MCP server safe?
- No scan can promise that. What its definitions show: 11 tools, 4 of which can change, send or delete something, and 2 findings from fixed checks. Allow the reads, ask before each change, and deny the deletes; the permissions block on this page does that.
Can the AgentsIndex MCP server delete data?
- No tool is classed as destructive: going by their definitions, none of them deletes, cancels or revokes anything.
Can the AgentsIndex MCP server do things that cannot be undone?
- Yes. 1 tool can send, pay, publish or merge: send_lead.
What tools does the AgentsIndex MCP server have?
- 11 tools: 1 irreversible, 3 write, 7 read. send_lead, submit_listing, create_order, my_listings, search_listings, get_listing, list_categories, compare_listings and 3 more.
Does the AgentsIndex MCP server need sign-in or OAuth?
- Its remote endpoint listed its tools to an anonymous client, without sign-in.
Does the AgentsIndex MCP server send data to outside services?
- 2 tools can carry data out (it sends, posts or publishes, or takes a recipient, channel or URL): submit_listing and send_lead.
Has the AgentsIndex MCP server changed its tools?
- Yes, 2 times since Anlyon Radar started tracking it, most recently on Nov 11, 2026. Every change is on this page and in its Atom feed.