Skip to content

AI Cortex MCP server

Which servers in the official MCP registry are AI Cortex's own, whether they answer, and what they offer. A program checked each of these, nobody copied them from a product page.

read 2026-10-26

← Back to AI Cortex

Whose server it is
  • Registry namedev.aic0rt3x.memory/mcpverified in the MCP registry

The registry name sits under aic0rt3x.dev. The registry only lets a publisher use that namespace after proving control of the domain through DNS or HTTP.

When we asked itchecked 2026-10-26
  • Stateanswerschecked 2026-10-26
  • Tools15
  • Server calls itselfai-cortex-storage
  • Server version1.0.0
  • Protocol version2025-03-26

On 2026-10-26 the server answered initialize and tools/list without a sign-in. First answer in 127 ms.

Should an agent connect itNot proven either way yet
  • Its tools do every use case we checked: keep an agent's notes available between separate sessions; look up saved memories from an MCP-compatible agent.
  • Checked on 1 of the 7 days it takes to judge reliability.

Once connected

  • For "keep an agent's notes available between separate sessions", the likely first call is put_memory.
  • For "look up saved memories from an MCP-compatible agent", the likely first call is list_namespaces.
  • delete_memory, create_api_key change data without a sign-in: anyone with the address can call them.

Said by rules from the measurements above. The likely first call is a model's pick from the tool list; read its description before relying on it.

Measured dailydaily checks: 1, the first on 2026-10-26
  • INITE MCP scoreafter a weeka score needs 7 daily checks: 1 so far
  • What an agent can do with it2 of 2 use cases
  • Answeredmeasuring1 of the 7 daily checks it takes to print a share
  • Median first answer203 ms
  • Space in an agent's context≈ 1,351 tokensthe definitions of all its tools (15), which an agent carries on every turn
  • Tool definitionsA · 94 of 100
  • Typed results0 of 15 toolsdeclared outputSchema: results come typed, read once
  • Not saying whether it only reads or can destroy: 5 of 15

What an agent can do with it

  • can do · keep an agent's notes available between separate sessions
  • can do · look up saved memories from an MCP-compatible agent
  • not sure · make one agent's memories reachable from several devices

Each of the product's use cases, put to Jev against this server's tool list: can an agent do it with these tools?

Worth a look

  • Tools that change data, on a server that answers without a sign-in: delete_memory, create_api_key

A program checks the server once a day: initialize and tools/list, no model. Tokens are the definitions' length at about four characters a token. The grade averages six checks of every tool's name, description, schema and annotations; it says how well the tools are described, not whether they work. The score weighs what was measured and nothing else: answering (30%), doing the product's job (25%), definitions (20%), context (10%) and use (15%). A part that was not measured is left out, and a hidden instruction caps the score at 40. The state of vendors' MCP servers: method and numbers →

Tools it offers15 tools · read 2026-10-26
  • get_memory

    Retrieve a stored value by namespace and key.

  • put_memory

    Store a value with optional tags, priority, and storage duration in months (default 1). value_json must be ≤ 1 MB. On overwrite, the existing expiry is scaled by the KB size ratio of old to new value.

  • delete_memory

    Delete a stored value.

  • list_keys

    List keys in a namespace with metadata (created_at, updated_at, updated_count, ttl, priority, last_modified_key_id). Returns up to 1000 keys sorted by recency. Expired keys are automatically excluded.

  • top_memories

    Return the top N highest-priority non-expired keys in a namespace, ordered by priority descending. Useful for loading the most important context when space is limited. Maximum n is 1000.

  • list_namespaces

    List all namespaces for this account.

  • search_by_tag

    Find (namespace, key) pairs with a given tag. Returns up to 1000 results.

  • get_account

    Get account status, credit balance, and available credits.

  • list_api_keys

    List all API keys on this account.

  • create_api_key

    Create a new live API key. Only available for active accounts. The key is shown once — store it immediately.

  • revoke_api_key

    Revoke an API key by ID. The key stops working immediately.

  • list_payments

    List payment history for this account.

  • settle_balance

    Charges the saved payment method to settle your outstanding balance. Requires explicit user approval before running. For trial accounts, returns the activation URL instead. Minimum charge: $5.

  • submit_exit_feedback

    Submit feedback when a trial ends without activation. Free to call — no credits charged. Only the first submission per account is recorded; duplicates are silently ignored. Call this when the user declines to activate or the trial is…

  • submit_feature_request

    Submit a feature request or product feedback. Charged at storage rate: ceil_kb(description_length) × 100 credits per KB. Available to trial and active accounts alike.

How to connect

Add the server to your MCP client config. Claude Desktop, Cursor and most other clients read this shape:

{
  "mcpServers": {
    "aic0rt3x": {
      "type": "http",
      "url": "https://memory.aic0rt3x.dev/mcp"
    }
  }
}

Add it

Claude Code, in a terminal

claude mcp add --transport http aic0rt3x https://memory.aic0rt3x.dev/mcp

Add to CursorAdd to VS Code

claude.ai, ChatGPT and other clients: add https://memory.aic0rt3x.dev/mcp as a custom connector in the settings.

Each line is something a program found when it went looking: the registry entry, the server's own answer, a spec that parses. A missing line means we did not find it, not that it does not exist.