agent-memory MCP server
dev.veritap/agent-memoryveritap.devstreamable-httpv0.3.0Scanned
Agent memory: free durable storage (any keypair, no signup) that survives sessions and machines.
12 tools: 2 do something that can't be undone, 7 change things and 3 only read.
- Reads private data · 0
- Reads text others write · 0
- Can send data out · 2
- No findings
Tools listed without sign-in. Remote posture
With Anlyon
2 tools here do something that can't be undone. Govern them.
- 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 agent-memory MCP server can do
The agent-memory MCP server is published by veritap.dev, a domain the MCP Registry verified. It has 12 tools. It can do things that cannot be undone, such as sending, paying or publishing (locker_send and locker_invite) and create or change data (7 tools: locker_capabilities, locker_nonce, locker_ack, locker_directory and 3 more), but cannot delete data. 3 tools only read.
2 tools can send data out of your system (locker_send and locker_invite).
None of the fixed checks matched anything in its tool definitions. That is not a safety guarantee.
Its remote endpoint lists its tools without sign-in.
Last scanned Nov 11, 2026, version 0.3.0.
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.
12 tools
Can't be undone 2Claude Code: ask
locker_inviteFree.Can send data outNo findings
Why it is in this group
- The name "locker_invite" says "invite": a word for sending, and what is sent cannot be taken back.
What it can reach
- outboundThe name "locker_invite" says "invite": it sends or publishes something to other people.
Description, as the server wrote it Free. Emits a one-paste onboarding kit for an agent on another vendor (Claude Code, Gemini CLI, Cursor, Codex, VS Code, ChatGPT, an AGENTS.md repo section, or generic). The same wallet key across your fleet = shared memory + mailbox: hand your operator the kit, they paste it, and your sibling agent can load what you save. Keys never touch this server — the kit carries a placeholder.
locker_sendDeliver data to any wallet-addressed mailbox — another agent's, or your own future self's.Can send data outNo findings
Why it is in this group
- The name "locker_send" says "send": a word for sending, and what is sent cannot be taken back.
- The input addresses people (input.to): what it sends cannot be taken back.
What it can reach
- outboundThe name "locker_send" says "send": it sends or publishes something to other people.
- outboundThe input takes a destination (input.to): what the tool writes goes there.
Description, as the server wrote it Deliver data to any wallet-addressed mailbox — another agent's, or your own future self's. PAID via x402: call without payment_b64 to get the price and accepts[] requirements, sign an EIP-3009 USDC authorization, retry with payment_b64. Bodies ≤32KB go inline (body_b64); larger declare body_upload with size_bytes and PUT to the returned upload_url. If the recipient registered require_e2e, the body must be sealed-box ciphertext for their locker_directory key, sent inline with encrypted:true. product:'receipt_vault' stores a flat-priced sealed receipt for 365 days.
Changes things 7Claude Code: ask
locker_ackFree, owner-signed.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.
Description, as the server wrote it Free, owner-signed. Ack IS delete — disclosed rule, by design: acked messages and their bodies are removed immediately. Unknown ids are silently skipped.
locker_capabilitiesFree.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.
Description, as the server wrote it Free. Returns the machine-readable contract — identity model, signing statements, prices, limits, custody commitments (deletion rules, 30-day sunset), x402 payment flow, and error codes. Call this first.
locker_checkpoint★ THE CORE PRODUCT.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.
Description, as the server wrote it ★ THE CORE PRODUCT. Durable memory for your agent, addressed by its wallet: store the state that must outlive this session, and a fresh process holding ONLY your wallet key loads it back byte-for-byte — no server to run, no second secret. A single agent gets full value alone; no other agent has to exist. Billed from prepaid credit (locker_credit) at the published GB-month rate. save: declare slot + size_bytes, PUT bytes to the returned upload_url (last 3 versions kept, 32 slots). load: returns a signed body_url (≤3 redemptions). list: all slots. delete: remove a slot. Credit exhaustion ⇒ 30 days read-only grace before expiry — top up to resume writes.
locker_creditPAID via x402 (same flow as locker_send: call without payment_b64 for requirements, retry with it).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.
Description, as the server wrote it PAID via x402 (same flow as locker_send: call without payment_b64 for requirements, retry with it). Prepaid credit funds checkpoint storage at the published GB-month rate; minimum $1, balance cap $100. Top-up clears read-only grace immediately.
locker_directoryFree.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.
Description, as the server wrote it Free. Returns the registered X25519 public key for an address plus the wallet-signed registration statement proving the key belongs to that wallet — verify the signature before sealing to it. 404 means no key registered (existence of mail is never leaked).
locker_nonceFree.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.
Description, as the server wrote it Free. Returns the single-use statement to sign (EIP-191) for authenticated tools. Expires in 5 minutes.
locker_register_keyOwner-signed.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.
Description, as the server wrote it Owner-signed. Publishes your X25519 public key so senders can seal to you. key_sig is your wallet's EIP-191 signature over 'veritap-locker:register-key:{address}:{enc_pubkey}' — the public proof the key is yours. Set require_e2e:true to make your mailbox reject non-ciphertext; private_count:true to make locker_count answer like an unused address. Recommended: derive the keypair from your wallet signature (see locker_capabilities identity.derived_enc_key) so your wallet stays your only secret.
Only reads 3Show 3 read-only tools
locker_countFree.No findings
Why it is in this group
- The name "locker_count" says "count": a word for reading.
Description, as the server wrote it Free. Count of unacked messages for an address — a cheap liveness/mail check that needs no signature. Addresses registered with private_count answer 0, indistinguishable from unused.
locker_readFree.No findings
Why it is in this group
- The name "locker_read" says "read": a word for reading.
Description, as the server wrote it Free. Returns unacked messages for your address — inline bodies as body_b64, large bodies as short-lived signed body_url (redeemable ≤3 times; re-read for a fresh link). Non-destructive: messages stay until you locker_ack them or their TTL expires. Paginate with next_cursor.
locker_statusFree.No findings
Why it is in this group
- The name "locker_status" says "status": a word for reading.
Description, as the server wrote it Free. Credit balance, stored checkpoint bytes, daily burn, projected empty date, and grace state for an address.
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.
- 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 out2
- 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, 18:53 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
- 9
- 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 9 tools that change things (2 irreversible, 7 write). Whether a call also works without credentials was not tested: tools are never called.
What this check looks forMatched text
https://locker.veritap.dev/mcp/memory: tools/list returned 12 tools to an anonymous client, including locker_send (irreversible), locker_invite (irreversible), locker_capabilities (write), locker_nonce (write), locker_ack (write) and 4 more
Findings
Matches of fixed checks, with the text they matched.
No check matched anything in these tool definitions. That is not a safety guarantee.
History
Changes to its tool definitions and to its remote endpoint, newest first.
No change seen yet. Its tools are rescanned every day, and a remote endpoint is looked at once a week.
Install
Pinned to version 0.3.0. Values in angle brackets are yours to fill in.
claude mcp add --transport http agent-memory https://locker.veritap.dev/mcp/memoryclaude mcp add --transport http agent-memory https://locker.veritap.dev/mcp/memoryRun 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__agent-memory__locker_read",
"mcp__agent-memory__locker_count",
"mcp__agent-memory__locker_status"
],
"ask": [
"mcp__agent-memory__locker_capabilities",
"mcp__agent-memory__locker_nonce",
"mcp__agent-memory__locker_send",
"mcp__agent-memory__locker_ack",
"mcp__agent-memory__locker_directory",
"mcp__agent-memory__locker_register_key",
"mcp__agent-memory__locker_checkpoint",
"mcp__agent-memory__locker_credit",
"mcp__agent-memory__locker_invite"
],
"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 agent-memory. Review it before you use it.
Capability tags behind 2 rules
- mcp__agent-memory__locker_send outbound
- mcp__agent-memory__locker_invite 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/dev.veritap/agent-memory/verify?hash=a5e88a46932eba7d8256438cc8ec87f532c499f8bc8f2e3c197f45c4623ad21d" | jq -e '.changed == false'Pinned hash a5e88a46932e, from Nov 11, 2026, 10:17 UTC.
Watch
Get told when its tools change: a new tool, a new description, a new class.
https://radar.anlyon.com/servers/dev.veritap/agent-memory/changes.xmlAn Atom feed. Works in any feed reader or Slack's RSS app.
Badge
[](https://radar.anlyon.com/servers/dev.veritap/agent-memory)Source: MCP Registry · repository. Synced Oct 21, 2026, 18:12 UTC.
agent-memory MCP server: questions and answers
Answered from the scan above, not from the publisher's description.
Is the agent-memory MCP server safe?
- No scan can promise that. What its definitions show: 12 tools, 9 of which can change, send or delete something, and 0 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 agent-memory MCP server delete data?
- No tool is classed as destructive: going by their definitions, none of them deletes, cancels or revokes anything.
Can the agent-memory MCP server do things that cannot be undone?
- Yes. 2 tools can send, pay, publish or merge: locker_send and locker_invite.
What tools does the agent-memory MCP server have?
- 12 tools: 2 irreversible, 7 write, 3 read. locker_send, locker_invite, locker_capabilities, locker_nonce, locker_ack, locker_directory, locker_register_key, locker_checkpoint and 4 more.
Does the agent-memory MCP server need sign-in or OAuth?
- Its remote endpoint listed its tools to an anonymous client, without sign-in.
Does the agent-memory 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): locker_send and locker_invite.
Has the agent-memory MCP server changed its tools?
- Not since Anlyon Radar started tracking it. Last scanned Nov 11, 2026.