Skip to content

README.md

Latest commit

 

History

History
50 lines (35 loc) · 1.36 KB

README.md

File metadata and controls

50 lines (35 loc) · 1.36 KB

Mnemon Harness

mnemon-harness is an experimental Agent Integration layer for connecting host agents to Local Mnemon.

The current product surface is intentionally small:

  • setup installs memory/skill integration assets into Codex or Claude Code.
  • local run starts the project-local Mnemon service.
  • status reports Agent Integration, Local Mnemon, and sync status.
  • sync connects Local Mnemon to a Remote Workspace (mnemon-hub) and pushes/pulls governed commits with attribution preserved.
  • loop validate remains hidden and is used by make harness-validate.

Host directories such as .codex and .claude are projection surfaces. Runtime state is under .mnemon/harness/, and release-path Mnemon behavior stays under cmd/ and internal/.

Build

From the repository root:

go build -o mnemon .
go build -o mnemon-harness ./harness/cmd/mnemon-harness

Validate harness declarations:

make harness-validate

Try The Harness

Install memory and skill integration for a host:

./mnemon-harness setup --host codex --loop memory --loop skill --project-root .
./mnemon-harness local run
./mnemon-harness status

Remove projected assets for a principal:

./mnemon-harness setup uninstall --host codex --loop memory --loop skill --principal codex@project --project-root .

More command examples are in docs/harness/USAGE.md.