DSH Memory Protocol
by Arypal
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| DSH_HOME | No | Host data directory. | ~/.dsh |
| MEMORY_MODEL_PATH | No | Directory containing the embedding model (must contain model.onnx). Auto-detected if unset; without it semantic search degrades to BM25. | |
| MEMORY_MOOD_STATE | No | Optional path to a JSON mood-state file used to weight memories. | |
| IMPORTANCE_API_KEY | No | API key for distill and LLM importance scoring (OpenAI-compatible, defaults to DeepSeek). | |
| MEMORY_SKILL_PYTHON | No | Python interpreter to use. Auto-detects python3, then python, then py. | |
| MEMORY_SKILL_DB_PATH | No | Path to the memory database. If explicitly set, it is always respected. | <DSH_HOME>/memory/memory.db |
| MEMORY_SKIP_BOOTSTRAP | No | Set to '1' to disable the missing-tool bootstrap self-check. | |
| MEMORY_IMPORTANCE_GATE | No | Set to 'llm' to have the LLM score importance on each ingest turn. | rule |
| MEMORY_PROTOCOL_BACKEND | No | Set to 'installed' to force use of the memory_skill package from site-packages for troubleshooting. |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Server capabilities have not been inspected yet.
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
No tools | |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
This server cannot be deployed
Maintenance
ActivityMaintained
ResponsivenessNo recent community issues