Hermes OKF: memory tool for Hermes Agent

Filesystem memory for Hermes Agent built on markdown and YAML bundles in Open Knowledge Format

What Hermes OKF does

Memory lives as plain .md and YAML files in an OKF bundle that you can cat, grep and keep in Git. An implicit knowledge graph is derived from markdown links between concepts, so there is no RDF or Cypher layer. The plugin implements the Hermes MemoryProvider interface and is registered with hermes-okf install-plugin, which creates ~/.hermes/plugins/hermes-okf and updates ~/.hermes/config.yaml to enable the plugin and set memory.provider.

Once installed, the hermes okf command supports search, list, show, snapshot and restore. A standalone hermes-okf CLI initializes bundles, validates conformance and runs validate-config, which performs 15 checks on the Hermes setup. Python decorators such as @memorize_decision and @memorize_tool record decisions and tool calls, bundles can be copied to another machine so an agent resumes from them, and an opt-in GitOKFBundle auto-commits on every session. The only core dependency is pyyaml, with RAG through LangChain and ChromaDB left optional.

Key features

Markdown and YAML memory bundles with no database
Native MemoryProvider plugin registered with hermes-okf install-plugin
hermes okf search, list, show, snapshot and restore commands
validate-config command with 15 checks on the plugin setup
Opt-in Git history that auto-commits per session
Optional RAG integration through LangChain and ChromaDB

When to use it

Keeping an auditable, diffable record of agent decisions
Moving an agent's memory to another machine by cloning the bundle
Snapshotting memory before a risky deployment and restoring it afterward

Who it is for: Hermes Agent users who want inspectable, version-controllable file memory instead of a database.

How it fits with Hermes Agent

It is built for Hermes Agent as a native memory provider plugin, selected through hermes memory setup.

How to install Hermes OKF

These commands are copied from the project's README. Check the repository for the latest steps before you run them.

pip install hermes-okf
hermes-okf install-plugin
hermes-okf validate-config
hermes memory setup

Requirements: pyyaml as the only core dependency; LangChain and ChromaDB are optional for RAG

FAQ

What is Hermes OKF?

Hermes OKF is a persistent memory system for Hermes Agent based on the Open Knowledge Format. It stores decisions, observations and project context as markdown with YAML frontmatter on your filesystem.

Does Hermes OKF work with Hermes Agent?

Yes, it is a native Hermes memory provider. After hermes-okf install-plugin, choose hermes-okf in hermes memory setup and use the hermes okf commands to search and inspect memory.

How do I install Hermes OKF?

Run pip install hermes-okf, then hermes-okf install-plugin and hermes-okf validate-config. Finish with hermes memory setup and select hermes-okf from the list.