saki-thinker
How it worksA brain and soul your agents can rent
saki-thinker is a hosted brain over MCP: describe a soul, issue a scoped key, and every runtime you ship reads standing context, deposits what it learns, and recalls it fenced to your workspace.
01 · Why
Deposits persist across sessions and survive a runtime swap — the same memory comes back fenced to your workspace, not scattered in one process.
Describe a soul once — identity, policies, beliefs — and every agent you connect inherits it. No re-prompting the character into existence.
Every read and write is stamped with its workspace and soul. Nothing your agent learns leaks sideways into another tenant's brain.
02 · Who it is for
03 · Setup
Give the agent its identity, policies and beliefs in the console.
A per-runtime credential with the exact read and write scopes it needs.
One MCP config line in your host — no SDK, no pasted secrets by hand.
The agent remembers what it learns and recalls it, workspace-fenced.
04 · How it works
get_context reads the soul's standing identity and policies.
remember writes one entry stamped with soul and workspace.
recall returns the right memories — nothing crossed the fence.
A hosted brain your runtime plugs into over one endpoint.
05 · Compatible with
One MCP config line connects Claude Code to the brain — scoped key, no SDK.
Hermes reads standing context and deposits memories, fenced to your workspace.
OpenClaw speaks to the hosted brain over the same single endpoint.
Anything that speaks MCP connects — the endpoint is the interface, not a vendor SDK.
06 · Start