Skip to main content
Glama

memory

Agent Memory: a memory your principal hands you and takes back at any time, the same for ChatGPT, Claude and Grok. Encrypted on your side; PazAIr stores only ciphertext and reads nothing. Your principal holds the memory key (pzm1_...): never send it. Derive from it the access token you send and the AES-256-GCM key you keep (HKDF-SHA256, format at /.well-known/agent-memory.json; client at /memory/client.js). Free: 100 entries; reading, exporting and erasing are always free. Pro: one payment for a year, no renewal, 10 000 entries and a Bitcoin proof per stamped entry (action upgrade). No key of PazAIr needed. Your principal sees and erases everything at /memory.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "action": "<action>",
      "access": "<access>"
    },
    "name": "memory"
  }
}

Input Schema

NameRequiredDescriptionDefault
idNoread, delete: the entry id (me_...)
afterNolist, export: the next value of the previous page
entryNowrite: {"v": 1, "alg": "A256GCM", "iv": "<12 bytes base64url>", "ct": "<ciphertext with tag base64url>"}, encrypted with the key derived with info "enc/v1"
limitNolist, export: entries per page, 1 to 100
stampNowrite, Pro: put this entry into the next daily root on Bitcoin and Stellar (proof in read)
accessYesthe 43-character access token derived from the memory key (HKDF-SHA256, info "access/v1"); never the memory key itself
actionYesopen (once, with principal_has_key), status, write (entry, optional stamp), read (id), list (after), export (after: entries with ciphertext, a page of 100), delete (id), erase (confirm), upgrade (a payment link for your principal)
confirmNoerase: true, after your principal agreed; it removes every entry for good
principal_has_keyNoopen: true only if your principal holds the memory key. A memory hidden from them is not allowed.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Added

TDQS

A3.5/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Beyond the sparse annotations, the description discloses substantial behavior: client-side AES-256-GCM encryption, ciphertext-only storage, principal-side revocation and visibility at /memory, tier quotas, and the absolute rule never to transmit the memory key. Gaps remain around what the destructive actions actually return, pagination behavior, and what happens when the free quota is exhausted. Note the tension with destructiveHint=false, since the definition elsewhere describes permanent erasure, but the description itself does not claim the tool is non-destructive.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness3/5

Is the description appropriately sized, front-loaded, and free of redundancy?

It is front-loaded with the resource definition and the security-critical rule, which is correct ordering. But the single dense paragraph mixes a marketing comparison, pricing model, and protocol details that could be trimmed into structured lines, and several clauses do not help an agent decide or invoke.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness3/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

This is a nine-action, cryptographically complex tool with no output schema, so the description carries a lot of weight. It covers encryption, key derivation, quotas and plan tiers well, but says nothing about what the actions return, how pages terminate, or how errors (bad key, quota exceeded) surface. Adequate for authentication, incomplete for operating across nine actions.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so the baseline is 3 and the schema already documents action, access, entry, confirm and the rest. The description reinforces the critical access-key derivation (HKDF-SHA256, info 'access/v1') and points to /.well-known/agent-memory.json and /memory/client.js for the exact format, which is genuinely additive. It adds little about entry, stamp, after or limit beyond what the schema states.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The opening sentence names the resource and its scope: an encrypted memory store the principal hands over and can revoke, with the write/read/erase capabilities implied. It is clearly distinguishable from the marketplace-oriented siblings. It loses a point because the framing is partly product pitch ('the same for ChatGPT, Claude and Grok', 'No key of PazAIr needed') rather than a crisp statement of the operation set.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines3/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Usage guidance is implied rather than stated: the free-tier ceiling of 100 entries and the Pro tier of 10 000 entries tell the agent when upgrade becomes relevant, and the key-handling warning tells it how to authenticate. There is no explicit when-to-use/when-not or routing to alternatives, and per-action selection is left entirely to the action enum's own schema text.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources