OpenClaw connector
Point OpenClaw at the remote MCP URL or REST API so it can discover a user's live One PKM scopes before any data moves.
MCP & AGENTS
Bring OpenClaw, Hermes Agent, or any external MCP agent to a user's One Personal Data. Discover available scopes, request explicit consent, then retrieve only the approved encrypted export. Trusted-device writes use a separate owner-authorized native path.
Use a self-serve developer token, OAuth authorization code with S256 PKCE, or an operations-provisioned partner service credential.
Ciphertext and wrapped-key metadata only; OpenClaw, Hermes, or your external connector decrypts client-side.
Scopes are discovered per user from PKM metadata instead of hardcoded globally.
The server stores encrypted personal data and enforces consent before access.
OPENCLAW / HERMES
OpenClaw and Hermes are first-class ways to connect outside agent infrastructure to One, but the protocol is not locked to either name. Any external MCP agent can follow the same steps: discover Personal Data scopes, ask for consent, wait for the user, then retrieve the approved encrypted export. Consent MCP does not grant write authority.
Point OpenClaw at the remote MCP URL or REST API so it can discover a user's live One PKM scopes before any data moves.
Use Hermes as the consent messenger: every external tool call becomes a clear request inside One with purpose, scope, expiry, and receipt.
If you are not using OpenClaw or Hermes, follow the same contract directly from your agent host: discover, request, wait, then retrieve the encrypted scoped export.
NOW ON CLAWHUB
The ClawHub package @hushh/one-mcp bundles the MCP config and OpenClaw skill guidance for Hussh One consent flows. It still requires a developer token and user-approved scopes before PKM data moves.
Mint a developer token and connect OpenClaw, Hermes Agent, or your own agent host through Remote MCP, REST, or the npm bridge.
Call search-user-scopes so the agent sees the user's current PKM scopes instead of guessing from a static menu.
Call request-consent with one discovered scope, a clear purpose, expiry, timeout, and connector key-wrapping metadata.
After approval, use the returned grant reference to retrieve the encrypted scoped export in your trusted connector process.
RUNTIME MODES
Pick the host shape first. Remote MCP is the cleanest path for agent hosts, REST is the direct application contract, and the npm bridge keeps stdio-only clients on the same protocol.
Authorization: Bearer <developer-token>Use a bare HTTPS MCP URL and keep the developer token machine-local in the host secret store.
/api/v1Use versioned endpoints for scope discovery, consent requests, status checks, and encrypted scoped exports.
npx -y @hushh/mcpUse the public launcher when an external agent still expects a local stdio process instead of remote MCP.
CONSENT FLOW
Call search-user-scopes with the person's identifier. Scopes are dynamic and come from their indexed PKM.
Ask for one discovered scope with a transparent purpose, expiry, timeout, and connector key-wrapping metadata.
The user approves or denies inside the hussh surface. Poll only with the returned request_ref and interval.
Use grant_ref and expected_scope to retrieve the encrypted scoped export in a trusted connector process. There is no plaintext fallback.
SCOPE MODEL
Always discover first. A broader active grant can satisfy a narrower request, but a narrower grant never silently upgrades to a broader parent scope.
attr.{domain_slug}.{scope_slug}.*A dynamic semantic branch returned by per-user discovery. Do not hardcode a domain key.
cap.one.invokeA reserved One task capability. It grants no user-data read or mutation authority by itself.
MCP TOOLS
The integration snippets and tool names below are read from the live consent API. Authentication identifies your app; every user-data request still requires an explicit, time-limited consent grant.