jevmem-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| JEVMEM_INDEX | No | Pluggable vector index backend, for example sqlite-vec, Qdrant, LanceDB or Postgres/pgvector. | |
| JEVMEM_SCOPE | No | Scope for notes. Defaults to global unless set, for example project:my-project. | |
| JEVMEM_EMBEDDER | No | Embedder to use for semantic search, for example fastembed for a local model with no API. | |
| TYPESAFE_API_KEY | Yes | Your TypeSafe API key for Jev. Read automatically from ~/.jevmem/.env if present. |
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 |
|---|---|
| memory_writeA | Store ONE literal fact (decision, bugfix, convention, gotcha, preference, event). Use explicit entity names and ABSOLUTE dates (never 'yesterday'): the memory model reads literally. Content that looks like instructions to an agent is rejected. scope: 'global' or e.g. 'project:'. pinned=True only when the user says this must always be remembered: pinned notes open every session. |
| memory_recallA | Adaptive recall. Check |
| memory_pinA | Pin (or unpin) a note the user wants in every session; SessionStart injects pinned notes first. |
| memory_listB | List stored memories (newest first) with their top memory types. |
| memory_forgetC | Delete a memory and its edges. |
| memory_pending_synthesisA | Merge/promote proposals from consolidation. YOU are the writer: for each item follow |
| memory_resolveB | Store your synthesized fact for a pending proposal. Source notes are kept (merged ones rank lower). |
| memory_dismissC | Reject a merge/promote proposal. |
| memory_consolidateA | Run a consolidation pass now (also runs automatically every 20 writes). all_notes=True re-judges every stored note from scratch (slow: one Jev call per note); use it once after upgrading jevmem. |
| memory_statsB | Node counts, queued (degraded) writes and Jev usage for this server process. |
| memory_flush_pendingB | Retry writes queued while Jev was unreachable. |
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 11 tools
Most tools have clearly distinct purposes: write, recall, list, forget, pin, stats, flush, and consolidation operations. There is minor overlap between memory_write with pinned=True and memory_pin, and the two 'pending' tools could momentarily confuse, but descriptions make the intended workflows clear.
Every tool uses a consistent snake_case memory_ prefix with readable verb or noun forms (memory_write, memory_recall, memory_pin, memory_forget, memory_consolidate). No naming deviations or mixed conventions.
11 tools is well-scoped for a persistent memory server that includes core CRUD, pinning, listing, stats, and an asynchronous consolidation/synthesis lifecycle. Each tool has a distinct role and none appears redundant.
The core memory lifecycle is well covered: write, recall, list, forget, pin, consolidate, resolve, dismiss, stats, and flush queued writes. A direct content-update tool for an existing memory is missing, so correcting a stored fact requires forgetting and rewriting, but this is a minor workaround gap.