project-memory-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| PROJECT_MEMORY_DB | No | Path to the SQLite database file. By default, the database is created at ~/.local/share/project-memory/store.sqlite (or under $XDG_DATA_HOME if set). | ~/.local/share/project-memory/store.sqlite |
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": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| memory_recordA | Record a durable memory (note, decision, requirement, or learning) for a project. Identical kind+title+body is stored once; the response reports created=false when it already existed. Pass supersedes= to mark an earlier memory as replaced by this one; superseded memories are hidden from search by default. |
| memory_searchA | Full-text search over a project's memories. Query words are matched literally (all words must appear) and the last word is a prefix match, so |
| project_statusA | Report whether a project has recorded memory, counts by kind, and the 5 latest entries. Read-only. |
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 has a clearly distinct purpose: recording, searching, and summarizing project memory. There is no overlap or ambiguity between them.
Names follow a predictable resource-prefixed pattern: memory_record, memory_search, project_status. The pattern is readable and consistent enough, though project_status is more of a noun phrase than an action-oriented name.
Three tools is well-scoped for a focused project-memory server. Each tool earns its place and the surface is small but not underwhelming.
Core memory workflows are covered: recording, searching, and checking status. A minor gap is the lack of a get-by-id or full-text retrieval tool, since search only returns excerpts; deletion is handled indirectly via supersede.