memtrust
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| ZEP_API_KEY | No | API key for Zep (required to run the zep backend). | |
| MEM0_API_KEY | No | API key for the Mem0 hosted platform (required to run the mem0 backend). | |
| OPENAI_API_KEY | No | OpenAI API key used by mem0_direct for embeddings and extraction. | |
| OPENVIKING_API_KEY | No | API key for OpenViking (required to run the openviking backend). | |
| MEMPALACE_STORAGE_PATH | No | Path to the MemPalace storage (required to run the mempalace backend). | |
| MEMTRUST_JUDGE_API_KEY | No | API key for the LLM judge used in LongMemEval/LoCoMo evals (needed for LLM-judged evals). | |
| MEM0_DIRECT_VECTOR_STORE_URL | No | URL for the vector store, default is http://localhost:6333. | |
| MEM0_DIRECT_EMBEDDER_PROVIDER | No | Embedder provider for the mem0_direct backend (e.g., openai). | |
| MEM0_DIRECT_VECTOR_STORE_PROVIDER | No | Vector store provider for the mem0_direct backend (e.g., qdrant). |
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
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": false
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| runA | Run the memtrust agent-memory benchmark harness and return the resulting JSON report. Real |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 1 tool
Only one tool exists, so there is no possibility of confusing it with another. The description clearly explains what `run` does.
The tool name 'run' is a simple verb without a noun, but since it is the only tool, there are no conflicting conventions. It follows a typical CLI command pattern.
With just a single tool, the server feels thin and offers only one entry point. However, given the server's narrow purpose of running a benchmark harness, the count might be justified.
The tool fully covers the server's purpose: running the memtrust benchmark with customizable args and handling missing backend credentials gracefully. No obvious missing operations.