Provenonce MCP server

by ProvenonceAI·io.github.ProvenonceAI/provenonce·v1.1.5

No reviews yet

Cryptographic identity, SIGIL naming, heartbeat, and passports for AI agents.

B74/100grade B

Write a review Install

What users say

No reviews yet

Be the first

Safety scan

B74/100

full report

Adoption

New

0 stars69 downloads/wk

Reviews

Write one

Nobody has reviewed Provenonce yet.

If you have run it, two minutes of your experience saves the next person an afternoon.

Provenonce tools (11, 4 write)

write = sends, deletes, buys or posts

Read from the package source without running it. The installed server may list more.

provenonce_batch_heartbeatwrite action
Submit heartbeats for multiple child agents in a single call. One payment covers all children. Requires active sponsorships for each child. Max 20 children per batch.
provenonce_beats_proof
Compute a Beats work-proof \u2014 cryptographic evidence of computational effort by this agent. Computes sequential hash chains at the current network difficulty, submits to the Beats service for a signed receipt. The receipt can be used for spawn authorization and other operations.
provenonce_get_lineage
Get the full lineage chain for an agent \u2014 all events in its provenance history (registration, SIGIL issuance, heartbeats, spawns). Public \u2014 no authentication required.
provenonce_get_passport
Get this agent's cryptographically signed Passport \u2014 a verifiable identity document any third party can verify offline using the Provenonce authority public key. Root agents (depth 0) require an active SIGIL; child agents (depth > 0) only need at least one heartbeat.
provenonce_heartbeatwrite action
Submit a heartbeat to prove this agent is alive. Root agents (depth 0) require a SIGIL first; child agents (depth > 0) can heartbeat without a SIGIL. If payment_tx is omitted, returns payment instructions (fee amount + address).
provenonce_purchase_sigilwrite action
Purchase a Provenonce SIGIL to unlock full provenance: heartbeat access, signed passport, and a permanent identity record. If payment_tx is omitted, returns payment instructions (fee amount + address). SIGIL format: name*principal*tier (e.g. my-agent*acme*ind).
provenonce_register
Register this agent with Provenonce to get a cryptographic identity (hash + API key). Safe to call multiple times \u2014 returns existing identity if already registered. No wallet required. Free.
provenonce_spawn
Spawn a child agent under this agent's identity. The child inherits lineage from the parent. Returns the child's hash and API key. Requires sufficient beats balance or a work-proof receipt.
provenonce_status
Get this agent's current Provenonce status: identity, SIGIL, beats balance, and next steps. Requires prior provenonce_register call.
provenonce_submit_beatswrite action
Compute VDF beats and submit to the Registry to credit this agent's lifetime beat count. Unlike provenonce_beats_proof (which gets a receipt for spawn/resync), this extends the agent's persistent beat chain and increases the "Lifetime Beats" metric visible in the Registry.
provenonce_verify_agent
Verify another agent's Provenonce identity. Returns their SIGIL, identity class, and registration status. Public \u2014 no authentication required.

Public scan report

scanner v0.1.10 · 2026-09-28 · same rubric, same numbers if you re-run it

no findings

Code scan7 source files scanned25/25
–Live reliabilityno gateway calls yet and no remote to proben/a
–Tool poisoningtools not inspected (local package is not executed); not countedn/a
Auth qualitylocal package, no credentials required12/15
Maintenancelast push 200 days ago4/15
Maintainer identityregistry namespace matches repository owner7/10

Overall 74/100. Components that don't apply are left out of the denominator. Any critical finding is an F.Rubric Appeal a finding JSON

How Provenonce compares

Against every graded MCP server, from our own scans and uptime checks. Updated after each scan run.

Safety score: 74/100; higher than 40% of 48,255 servers
Tools: 11; the typical server has 7
Tools that change things: 36%; typical: 0%. Fewer write tools means less an agent can break
Downloads a week: 69; more than 53% of 12,191 servers with a package

What its tools let an agent do: 7 read · 4 change

Install Provenonce in Claude Code, Cursor or VS Code

Runs npx -y @provenonce/mcp on your machine. Read the scan report first; the gateway never runs local packages.

claude mcp add provenonce -- npx -y @provenonce/mcp

Add to Cursor

What the publisher says

From the Provenonce repository's README, as published. We do not edit it. Read it on GitHub

@provenonce/mcp

Provenonce Skill — Give any AI agent cryptographic identity in under 60 seconds.

Add this skill to your agent and it gets:

A permanent cryptographic hash (identity)
A SIGIL (verified identity class + tier)
A signed Passport (verifiable offline by any third party)
Heartbeat (continuous liveness proof)
Beats proofs (evidence of computational work)
Lineage (full ancestry chain)

Quick Start

Claude Desktop / Claude Code

Add to ~/.claude/claudedesktopconfig.json:

{
  "mcpServers": {
    "provenonce": {
      "command": "npx",
      "args": ["-y", "@provenonce/mcp"]
    }
  }
}

That's it. On first use, your agent calls provenonce_register and gets a permanent identity.

With pre-existing credentials

{
  "mcpServers": {
    "provenonce": {
      "command": "npx",
      "args": ["-y", "@provenonce/mcp"],
      "env": {
        "PROVENONCE_API_KEY": "pvn_...",
        "PROVENONCE_AGENT_HASH": "0x..."
      }
    }
  }
}

As a skill developer (earn rev share)

If you're a developer embedding this skill in your framework, include your partner ref token:

{
  "env": {
    "PROVENONCE_SKILL_REF": "your16hextoken"
  }
}

You earn 30% of every SIGIL purchased by agents you deploy. Register at POST /api/v1/skill/register.

The 11 Tools

Usage Examples

Example 1: Register and check status

User prompt: "Register with Provenonce and show me my identity."

Expected behavior:

Agent calls provenonce_register with an optional name
Receives a permanent cryptographic hash (e.g. 0xabcd...) and API key
Credentials are saved to ~/.provenonce/config.json
Agent calls provenonce_status to display identity, balance, and next steps
Status response includes a _hint suggesting the agent purchase a SIGIL

Example 2: Verify another agent before trusting it

User prompt: "Check if agent 0xf7dad8f4...b6fd is legitimate."

Expected behavior:

Agent calls provenonceverifyagent with the target hash
Returns the agent's SIGIL, identity class, tier, registration date, and liveness score
Agent reports whether the target has an active SIGIL and recent heartbeats
No authentication needed — this is a public verification endpoint

Example 3: Full lifecycle — SIGIL, heartbeat, and passport

User prompt: "Get me a SIGIL, send a heartbeat, and export my passport."

Expected behavior:

Agent calls provenoncepurchasesigil without payment_tx to get fee instructions (amount in SOL + ops wallet address)
User sends the SOL payment and provides the transaction signature
Agent calls provenoncepurchasesigil again with the payment_tx to complete the purchase
Agent calls provenonce_heartbeat (same payment flow) to prove liveness
Agent calls provenoncegetpassport to receive a cryptographically signed identity document
Passport can be verified offline by any third party using the Provenonce authority public key

Agent Flow

Shortened. The full README is on GitHub.

Nothing above is checked by us. What we check is on the safety report.

Provenonce: common questions

Is Provenonce MCP server safe?: Mostly: it is graded B (74/100). Read the Provenonce safety report
How do I install Provenonce?: It runs on your machine. Copy the Claude Code, Cursor, VS Code or Claude Desktop config from the install section.
Does Provenonce need an API key?: Not as far as the registry entry and our scan can tell: no credentials are declared or required.
Is Provenonce maintained?: The last commit was 237 days ago (2026-03-12). The latest release is v1.1.5.
How does Provenonce compare with other MCP servers?: safety score 74/100 (higher than 40% of 48,255 servers); tools 11 (the typical server has 7); tools that change things 36% (typical: 0%. Fewer write tools means less an agent can break). Full comparison.