Mnemosure
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| DASHSCOPE_API_KEY | Yes | Your Qwen Cloud (DashScope) API key. Required at runtime. | |
| MNEMOSURE_DATA_DIR | No | Override the directory where memories are stored (default: ~/.mnemosure). | |
| MNEMOSURE_MODEL_BRAIN | No | Override the main answer model ID (default: qwen3.7-plus-2026-05-26). | |
| MNEMOSURE_MODEL_EMBED | No | Override the embedding model ID (default: text-embedding-v4). | |
| MNEMOSURE_MODEL_RERANK | No | Override the rerank model ID (default: qwen3-rerank). | |
| MNEMOSURE_MODEL_CLASSIFY | No | Override the extraction/judge/classify model ID (default: qwen3.5-flash). |
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 |
|---|---|
| recallA | Answer a question grounded in stored memories. Says "not in the record" when it doesn't know; cites sources when it does; corrects superseded facts. query: the question (any language — the answer follows the question's language) returns: {"confidence": "certain|vague|unknown", "answer": grounded answer with sources, "cited": [memory ids]} |
| rememberA | Extract durable memories (decisions, changes, failures, facts) from a conversation or session text. When a new decision overrides an old one, links supersedes/because automatically. session_text: the raw conversation/session text to remember date : 'YYYY-MM-DD' (defaults to today) title : human-readable session title used for source citations (optional) returns : {"stored": [{"id", "content", "kind"}, ...], "count": n} |
| list_memoriesA | List the memory warehouse (by default only currently-valid memories). include_superseded: when True, also include old memories that have been superseded each item: {"id", "content", "kind", "scope", "status", "source"} |
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 3 tools
Each tool serves a distinct purpose: listing memories, answering questions from memories, and storing new memories. There is no functional overlap.
All tool names follow a consistent verb or verb_noun pattern in lowercase with underscores: list_memories, recall, remember.
Three tools cover the core memory operations (list, query, store) well, but a few more (e.g., delete, update) might be expected for full lifecycle management.
The tool set lacks explicit delete and update operations. While remember can supersede old memories, there is no way to independently modify or remove memories.