CF Memory Plugin

Persistent memory for AI agents, backed by Cloudflare Agent Memory.

This plugin is a thin client. Cloudflare stores, classifies, and recalls memories. You do not run a vector DB, embeddings pipeline, or Worker.

Private beta. Paid Workers is required. Paid Workers alone is not enough — you still need Agent Memory entitlement.

Need	Link
Join the beta	Waitlist form
Product docs	developers.cloudflare.com/agent-memory
HTTP API	HTTP API
Pricing	Agent Memory pricing
Limits	Platform limits
Create a token	API Tokens
Create a token (docs)	Create API token
Workers Paid	Workers pricing
Design notes	Introducing Agent Memory

60-second setup

pip install git+https://github.com/hansakoch/cf-memory-plugin.git

export MCP_CLOUDFLARE_API_KEY="cf-api-token-with-agent-memory"
export CF_ACCOUNT_ID="your-32-char-account-id"

cf-memory test

CF_ACCOUNT_ID is required. There is no default account. Find it in the Cloudflare dashboard sidebar.

Create the token at dash.cloudflare.com/profile/api-tokens with Agent Memory permission only. Do not reuse a Global API Key.

What this is

You call	Cloudflare does
remember	Store one fact / instruction / event
recall	Search + synthesize an answer (~5s)
ingest	Extract memories from a conversation (writes land 3–8s later)
summary	Markdown profile of what is stored
list / get / delete	Inspect or remove entries

Cost

Item	Today	Source
Agent Memory	$0 during private beta. 30-day notice before billing.	Pricing
Workers Paid (required to apply)	$5/month minimum	Workers pricing
This plugin	Free (MIT). Your own HTTP calls only.	—

Recommended: Workers Paid on the account that will hold memory. Do not put production memory on a free account. Do not assume every paid Workers account has Agent Memory — we verified paid ≠ entitlement.

After beta, treat Agent Memory as a separate bill. Cloudflare has not published GA rates yet.

Dependencies

Runtime:

Package	Why
httpx	HTTPS client to api.cloudflare.com
mcp	MCP server (cf-memory serve)

Optional:

Extra	Packages	When
pip install 'cf-memory-plugin[a2a]'	starlette, uvicorn	A2A peer server
pip install 'cf-memory-plugin[dev]'	pytest, pytest-asyncio, respx	Tests

No Cloudflare Worker, Wrangler, D1, Vectorize, or Workers AI binding is required for this plugin. Those are Cloudflare products this client does not use.

Security

Follow Cloudflare's token rules: API token best practices.

• Store MCP_CLOUDFLARE_API_KEY in the environment or a secret store. Never commit it.
• Scope the token to Agent Memory on one account.
• Set CF_ACCOUNT_ID yourself. This plugin will not fall back to another account.
• cf-memory a2a binds to 127.0.0.1 by default. Do not expose it to the public internet without auth.
• ingest and sync_turn send conversation text to Cloudflare. Do not ingest secrets, passwords, or customer PII you are not allowed to store.
• Use one namespace per app and one profile per user/tenant.
• Rotate the token from the API Tokens page if it leaks.

See SECURITY.md.

Use it from an agent

Same MCP block everywhere. Only the config file path changes.

{
  "mcpServers": {
    "cf-memory": {
      "command": "cf-memory",
      "args": ["serve"],
      "env": {
        "MCP_CLOUDFLARE_API_KEY": "your-token",
        "CF_ACCOUNT_ID": "your-account-id"
      }
    }
  }
}

Agent	Where to put it	Docs
Claude Code	.claude/mcp.json or claude mcp add cf-memory -- cf-memory serve	MCP
Codex	~/.codex/config.toml → [mcp_servers.cf-memory]	OpenAI Codex
Cursor	.cursor/mcp.json	Cursor MCP
Hermes	hermes config set memory.provider cloudflare-memory	Memory providers
OpenClaw	host MCP config	OpenClaw
TRAE	.trae/mcp.json	TRAE
OpenCode	~/.opencode/config.json	OpenCode
pi	host MCP config	pi
Any MCP client	stdio cf-memory serve	MCP spec
LangChain / LangGraph	Python CloudflareMemoryClient	LangChain
A2A peers	cf-memory a2a --port 9120	Agent card at /.well-known/agent.json

Hermes is the only native provider. It prefetches in the background so recall does not add ~5s to every turn.

hermes config set memory.provider cloudflare-memory
hermes cloudflare-memory test

Python:

import asyncio
from cloudflare_memory import CloudflareMemoryClient

async def main():
    async with CloudflareMemoryClient(
        account_id="your-account-id",
        api_token="your-token",
        namespace="my-app",
        profile="default",
    ) as client:
        await client.remember("User prefers concise answers.")
        result = await client.recall("How should I answer?")
        print(result.answer)

asyncio.run(main())

CLI

cf-memory test                          # connectivity
cf-memory serve                         # MCP (stdio)
cf-memory a2a --port 9120               # A2A on localhost
cf-memory card                          # print agent card

hermes cloudflare-memory status
hermes cloudflare-memory namespaces

Deploy the landing page (optional)

The plugin itself is not a Worker. The public page at cloudflare-memory.pages.dev is a static Cloudflare Pages site.

# token needs Pages:Edit on the same account
export CLOUDFLARE_API_TOKEN=...
export CLOUDFLARE_ACCOUNT_ID=...
npx wrangler pages deploy ./public --project-name cf-memory-plugin

Pages docs · Wrangler

Credits

• Memory backend: Cloudflare Agent Memory (blog, Discord, Community)
• Protocol: Model Context Protocol
• Hermes provider contract: Nous Research Hermes Agent
• HTTP client: httpx

This repo is not affiliated with Cloudflare or Nous Research.

License

MIT