Verify post-quantum (ML-DSA-65 / NIST FIPS 204) signatures and buy proof, provenance and AI-decision receipts over x402 (USDC on Base). MCP server. Public source mirror of the npm package. Not a FIPS-validated CMVP module.
- ✓Open-source license (MIT)
- ✓Actively maintained (<30d)
- ✓Clear description
- ✓Topics declared
- ✓Documented (README)
git clone https://github.com/johnInarti/fractalai-mcp-server{
"mcpServers": {
"fractalai": {
"command": "node",
"args": ["/path/to/fractalai-mcp-server/dist/index.js"]
}
}
}MCP Servers overview
# FractalAI MCP Server
A [Model Context Protocol](https://modelcontextprotocol.io) (MCP) server that exposes
FractalAI's live **post-quantum cryptography** and **x402 micropayment** capabilities as
tools any MCP client (Claude Desktop, IDEs, agent frameworks) can call over stdio.
It talks to the real, public FractalAI deployment at `https://fractalai.net.co` — nothing
here is mocked.
## What it does
| Tool | Cost | What it does |
|------|------|--------------|
| `verify_proof` | **free** | Verify a post-quantum **ML-DSA-65 / Dilithium-3 / NIST FIPS 204** signature against its public key. Returns `{ valid, verdict }`. |
| `x402_checkout` | **free** | Return live [x402](https://x402.org) (HTTP 402) payment requirements for a paid resource — price, network, USDC asset, `payTo`, and the `accepts[]` entries. Also fetches the full `.well-known/x402.json` manifest. |
| `pqc_sign` | **$0.02 USDC** | Buy one ML-DSA-65 / Dilithium-3 signature over a `message` (its sha256 is signed) or a hex `hash` (signed directly). |
| `seal_provenance` | **$0.05 USDC** | Buy a PQC + [drand](https://drand.love) provenance seal over content: a drand round + a self-contained ML-DSA-65 / Dilithium-3 signature + an on-chain anchor attempt (this server's node-bound Dilithium-2 key) over the content hash. |
| `attest_decision` | **$0.05 USDC** | Seal an AI agent's declared decision (input hash, output, model, timestamp) — chained in an append-only Black Box, anchored on-chain (VAID-1, node-bound Dilithium-2 key), and returned with a self-contained ML-DSA-65 / Dilithium-3 signature for independent offline verification. |
## Honest scope (read this)
- **The signature every tool here serves/verifies directly is ML-DSA-65 / Dilithium-3 (NIST
FIPS-204, Cat-3).** `seal_provenance` and `attest_decision` ALSO write a separate on-chain
anchor (VAID-1) using this server's node-bound Dilithium-2 key — additive, not the signature
returned to you.
- **Dilithium-3 is resistant to known classical and quantum attacks per NIST FIPS 204 — it is
NOT "unbreakable".** A signature proves the *exact bytes* you submitted were signed by this
server's post-quantum key. It says **nothing** about the truthfulness of the signed content.
- `seal_provenance` produces a **provenance record**, not a content audit.
- `attest_decision` **notarizes a decision as declared**; it does not judge whether the decision
was correct.
- The paid tools run a **real HTTP 402 flow**. Without a `payment_tx_hash` they return the
payment *challenge* (`accepts[]`) — they never fabricate a result.
## Payment model (x402)
Paid tools are gated by [x402](https://x402.org): pay in **USDC on Base** to the service's
treasury, then the tool returns the resource. This server uses the **facilitator-less txHash**
settlement path, which needs no facilitator account:
1. Call the tool once with no `payment_tx_hash`. You get back the `accepts[]` block: the exact
USDC amount, the `asset` (USDC on Base), and the `payTo` treasury address.
2. Send that USDC transfer on Base yourself.
3. Call the tool again with `payment_tx_hash` set to the confirmed transaction hash. One
txHash buys exactly one call (anti-replay is enforced server-side).
`verify_proof` and `x402_checkout` are always free.
## Install & build
```bash
cd integrations/mcp
npm install
npm run build
```
## Configure in an MCP client
### Claude Desktop (`claude_desktop_config.json`)
```json
{
"mcpServers": {
"fractalai": {
"command": "node",
"args": ["/absolute/path/to/integrations/mcp/dist/index.js"]
}
}
}
```
### Any stdio MCP client
Run the built entrypoint: `node dist/index.js`. It speaks MCP over stdio; all logging goes to
stderr so stdout stays a clean JSON-RPC channel.
### Environment
- `FRACTALAI_BASE_URL` — override the endpoint base (defaults to `https://fractalai.net.co`).
## Example flows
**Verify a signature (free):**
```json
{
"name": "verify_proof",
"arguments": {
"signed_message_b64": "<base64 of the exact bytes that were signed>",
"signature": "<base64 ML-DSA-65 / Dilithium-3 signature>",
"public_key": "<base64 ML-DSA-65 / Dilithium-3 public key>"
}
}
```
**Price a signature (free), then buy it (paid):**
```json
{ "name": "x402_checkout", "arguments": { "resource": "sign" } }
```
```json
{ "name": "pqc_sign", "arguments": { "message": "hello world" } }
```
The first `pqc_sign` call (no `payment_tx_hash`) returns the payment challenge. Pay the USDC,
then call again:
```json
{
"name": "pqc_sign",
"arguments": { "message": "hello world", "payment_tx_hash": "0x…" }
}
```
## About FractalAI
FractalAI is an AI-native Layer 1 blockchain (chain id 62124) written in Rust with
post-quantum cryptography (ML-DSA-65 / Dilithium-3 signatures, Kyber encryption). These x402
endpoints sell post-quantum proof, provenance and decision-attestation services to autonomous
agents. See `https://fractalai.net.co` and the discovery manifest at
`https://fractalai.net.co/.well-known/x402.json`.
## License
MIT
What people ask about fractalai-mcp-server
What is johnInarti/fractalai-mcp-server?
+
johnInarti/fractalai-mcp-server is mcp servers for the Claude AI ecosystem. Verify post-quantum (ML-DSA-65 / NIST FIPS 204) signatures and buy proof, provenance and AI-decision receipts over x402 (USDC on Base). MCP server. Public source mirror of the npm package. Not a FIPS-validated CMVP module. It has 0 GitHub stars and its last recorded update is dated 2026-10-09.
How do I install fractalai-mcp-server?
+
You can install fractalai-mcp-server by cloning the repository (https://github.com/johnInarti/fractalai-mcp-server) or following the README instructions on GitHub. ClaudeWave also provides quick install blocks on this page.
Is johnInarti/fractalai-mcp-server safe to use?
+
Our security agent has analyzed johnInarti/fractalai-mcp-server and assigned a Trust Score of 95/100 (tier: Verified). See the full breakdown of passed checks and flags on this page.
Who maintains johnInarti/fractalai-mcp-server?
+
johnInarti/fractalai-mcp-server is maintained by johnInarti. The last recorded GitHub activity is dated 2026-10-09, with 0 open issues.
Are there alternatives to fractalai-mcp-server?
+
Yes. On ClaudeWave you can browse similar mcp servers at /categories/mcp, sorted by popularity or recent activity.
Deploy fractalai-mcp-server to your cloud
Ship this repo to production in minutes. Each platform spins up its own environment with editable env vars.
Maintain this repo? Add a badge to your README
Drop the badge into your GitHub README to show it's tracked on ClaudeWave. Each badge links back to this page and reflects the live Trust Score.
[](https://claudewave.com/repo/johninarti-fractalai-mcp-server)<a href="https://claudewave.com/repo/johninarti-fractalai-mcp-server"><img src="https://claudewave.com/api/badge/johninarti-fractalai-mcp-server" alt="Featured on ClaudeWave: johnInarti/fractalai-mcp-server" width="320" height="64" /></a>More MCP Servers
Fair-code workflow automation platform with native AI capabilities. Combine visual building with custom code, self-host or cloud, 400+ integrations.
User-friendly AI Interface (Supports Ollama, OpenAI API, ...)
An open-source AI agent that brings the power of Gemini directly into your terminal.
Real-time global intelligence dashboard. AI-powered news aggregation, geopolitical monitoring, and infrastructure tracking in a unified situational awareness interface
🕷️ An adaptive Web Scraping framework that handles everything from a single request to a full-scale crawl! Don't be shy, join here: https://discord.gg/EMgGbDceNQ and follow here for daily tips and tricks: https://x.com/Scrapling_dev
The fastest path to AI-powered full stack observability, even for lean teams.