OpenClaw + Hermes Agent
Connect OpenClaw or Hermes to a user's One PKM. External agents can request scoped context, import the approved data, and, when the person allows it, write updates back with a receipt.
One PKM consent rail
Let agents use PKM data without owning it.
External agent or MCP server
An agent asks for PKM context or permission to update it.
OpenClaw / Hermes connector
Route the request to One through the same consented MCP rail.
One PKM consent
The user sees the purpose, scope, expiry, and read/write access.
Scoped import or update
Approved context leaves One, or approved changes write back.
New request in One
An external agent wants to import selected PKM context and save a result back to One. Review the purpose, expiry, and receipt before anything moves.
Watch the demo.
Learn how to set up OpenClaw or Hermes with One PKM, show external agents what they can request, and turn every import or PKM update into a user-approved flow.
Any agent can request PKM access. One decides.
Connect OpenClaw or Hermes to the user's One PKM so external agents can import the exact context they are allowed to use, and write back only when the user approves it.
Users choose which PKM data an external agent can import.
Read and write scopes are explicit before access is granted.
Approved agents can update the PKM only within the accepted scope.
Every approval, denial, expiry, revocation, and writeback leaves an audit trail.
One consent inbox
Pending access requests
Portfolio agent
Import approved identity, preferences, and profile memory from One PKM
Research agent
Read selected notes and saved sources for a project brief
Ops agent
Write a completed task summary back to the user's PKM
One consent bridge for OpenClaw or Hermes.
Use OpenClaw, Hermes, or both as the same consented bridge between external MCP agents and the user's private One PKM. The point is controlled data movement: scoped import, approved writeback, and a clear audit trail.
Scoped PKM import
External agents can request selected memories, preferences, profile fields, notes, sources, tasks, or receipts from One. The user approves exactly what leaves the PKM, why, and for how long.
Consented PKM writeback
When the user allows it, agents can update One after the job is done: save new facts, refresh preferences, append research, complete tasks, or attach receipts with audit history.
Built for agents that need personal context.
Whether the agent runs in a desktop app, an enterprise workflow, a custom MCP host, or a model-native tool call, OpenClaw/Hermes gives it a clean way to ask One for PKM context and return approved updates.
Agent platforms
Sell agents that feel personal without asking users to hand over their whole PKM.
Enterprise workflows
Let employees approve project-specific context and writebacks instead of copying private data into another admin surface.
Personal automations
Give finance, travel, learning, and productivity agents a clean way to request context and save outcomes.
Developer tools
Use the technical MCP surface in /developers, then send buyers here to understand how One keeps PKM data permissioned.
Sell agents that connect to One PKM with consent.
Show users the promise here: their PKM can power outside agents without becoming a data free-for-all. Send builders to /developers when they are ready for MCP tools, scopes, consent requests, and implementation details.