three.ws Provenance runs on your own machine — the client starts it, so there is no endpoint to ping. 79 installs a week from npm. Last commit 1 Nov 2026.
Append-only, signed, on-chain-verifiable agent action log — record and audit what agents did.
Today is the operative word: we check three.ws Provenance every 15 minutes and re-read its code on every release. Watch it and you find out the day that stops being true.
This server runs on your own machine — install it with the package manager and the client starts it for you. Package name taken from the official registry entry.
claude mcp add provenance-mcp -e 'THREE_WS_TOKEN=<THREE_WS_TOKEN>' -- npx -y @three-ws/provenance-mcp
{
"mcpServers": {
"provenance-mcp": {
"command": "npx",
"args": [
"-y",
"@three-ws/provenance-mcp"
],
"env": {
"THREE_WS_TOKEN": "<THREE_WS_TOKEN>"
}
}
}
}
[mcp_servers.provenance-mcp]
command = "npx"
args = ["-y", "@three-ws/provenance-mcp"]
env = { THREE_WS_TOKEN = "<THREE_WS_TOKEN>" }
{
"mcpServers": {
"provenance-mcp": {
"command": "npx",
"args": [
"-y",
"@three-ws/provenance-mcp"
],
"env": {
"THREE_WS_TOKEN": "<THREE_WS_TOKEN>"
}
}
}
}
{
"servers": {
"provenance-mcp": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@three-ws/provenance-mcp"
],
"env": {
"THREE_WS_TOKEN": "<THREE_WS_TOKEN>"
}
}
}
}
Put your own value in place of
<THREE_WS_TOKEN> (Bearer token authenticating every call — a three.ws API key (sk_live_…) or an OAuth access token. The ledger is owner-scoped: you can read and append actions only for agents your token owns. Required for all three tools. Treat like a password).
The author’s registry entry marks it as required.
Optional settings from the registry entry, left out of the block above:
THREE_WS_BASE (Base URL of the three.ws API that serves the agent-actions ledger. Override only when self-hosting or targeting a preview deployment), THREE_WS_TIMEOUT_MS (Per-request timeout in milliseconds for the ledger reads/appends), THREE_WS_SIGNER_KEY (Optional 0x-prefixed EVM private key. When set, append_agent_action ERC-191-signs each action so anyone can ecrecover the signer and verify authorship on-chain. A per-call `signer_key` arg overrides it. Omit to append unsigned. Treat like cash).
Verifiable action logs for AI agents. Every action is recorded in a SHA-256 hash chain.
Bitcoin-anchored, tamper-evident audit log for AI agents — record, disclose and verify actions.
Verifiable cognition: prove a session ledger is tamper-evident; audit & import any agent's logs.
Read, verify and export a tamper-evident log of what an AI agent did. Read-only.
Human approvals, autonomy limits and a chained audit log for the AI agents of service firms.
Tamper-evident evidence log: signed heads, proofs, and recording salted digests of your records.
Verifiable crypto signals, recorded history and always-on monitoring for AI agents via x402 on Base.
Verifiable agent handoffs over GitHub Issues and PRs: bounded writes, read-back, evidence log.
Answers built from our own checks of this server.