mem0-lite
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| MEM0_DIR | No | Data directory for mem0-lite. | ~/.mem0 |
| OPENAI_API_KEY | No | Required unless using Ollama. Your OpenAI API key. | |
| MEM0_LITE_DEBUG | No | Set to enable debug logging to debug.jsonl. | |
| MEM0_LITE_GIT_GH | No | Background git GitHub integration. Default off. | off |
| MEM0_LITE_USER_ID | No | Default user scope. | $USER |
| MEM0_LITE_AGENT_ID | No | Optional agent scope. | |
| MEM0_LITE_GIT_FETCH | No | Background git fetch. Default off. | off |
| MEM0_LITE_GIT_ROOTS | No | Pathsep-separated allowlist for git toplevels. | |
| MEM0_LITE_LLM_MODEL | No | LLM model name. Defaults to llama3.2 with Ollama. | llama3.2 |
| MEM0_LITE_GIT_PROMOTE | No | Background git promote. Default off. | off |
| MEM0_LITE_LLM_PROVIDER | No | LLM provider, e.g. ollama. Default is OpenAI via mem0ai. | |
| MEM0_LITE_LOCK_TIMEOUT | No | Store lock wait in seconds. | 30 |
| MEM0_LITE_FEEDBACK_MODE | No | Set to '1' to enable feedback mode. Default off. | off |
| MEM0_LITE_EMBEDDER_MODEL | No | Embedder model. Defaults to nomic-embed-text with Ollama. | nomic-embed-text |
| MEM0_LITE_PLUGINS_DISABLE | No | Comma-separated plugin names to skip, e.g. git. | |
| MEM0_LITE_EMBEDDER_PROVIDER | No | Embedder provider, e.g. openai or ollama. |
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 |
|---|---|
| add_memoryA | Store a memory. Prefer infer=false with a self-contained third-person fact. |
| search_memoriesA | Semantic search. Rewrite the query to 3-6 keywords matching stored third-person facts. |
| get_memory_by_idA | Fetch one memory by id. When MEM0_LITE_FEEDBACK_MODE is enabled, rate via rate_memory_call(call_ts). |
| list_memoriesC | List memories in scope. Not a substitute for search_memories. When MEM0_LITE_FEEDBACK_MODE is enabled, responses include ts; rate useful/miss/noise via rate_memory_call. |
| update_memoryA | Replace memory text in place. Prefer this over delete+add when a fact changed. |
| delete_memoryC | Delete one memory by id. |
| delete_all_memoriesC | Delete all memories in scope. Requires confirm=true. |
| rate_memory_callA | Rate a prior retrieval by call_ts (from response ts when MEM0_LITE_FEEDBACK_MODE is enabled). |
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 8 tools
Each tool has a clearly distinct purpose: add, update, search, get by ID, list, delete single, delete all, and rate. Potential confusion between search_memories and list_memories is addressed in descriptions, ensuring no ambiguity.
All tool names follow a consistent verb_noun pattern in snake_case (e.g., update_memory, search_memories, delete_all_memories). The naming is predictable and uniform across the set.
With 8 tools, the server is well-scoped for a memory management system. The count covers essential operations without being excessive or insufficient.
The tool surface provides complete coverage for memory CRUD (add, get, update, delete) plus search, listing, bulk deletion, and a feedback rating mechanism. No obvious gaps for the declared purpose.