session-memory
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| LOG_LEVEL | No | Log level (error, warn, info, debug, trace). | info |
| SESSION_DB | No | Alternative database file path for the MCP server as used in client configuration. | ~/.agents/memory/session.db |
| ENABLE_CORS | No | Enable Cross-Origin Resource Sharing. | true |
| OPEN_BROWSER | No | Auto-open browser on startup. | false |
| DASHBOARD_HOST | No | Server host for the web dashboard (use 0.0.0.0 for all interfaces). | localhost |
| DASHBOARD_PORT | No | Server port for the web dashboard. | 3001 |
| SESSION_DB_PATH | No | Database file path for the session memory. | ~/.agents/memory/session.db |
| MCP_DASHBOARD_TOKEN | No | Bearer token for API authentication (leave empty to disable authentication). |
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 | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| store_session_contextB | Store session context for later retrieval |
| retrieve_session_contextC | Retrieve stored session context |
| update_session_contextA | Append to existing session context without overwriting |
| track_user_preferenceC | Track and learn user preferences |
| get_user_preferencesC | Retrieve user preferences |
| learn_project_conventionC | Learn and store project-specific conventions |
| get_project_conventionsC | Retrieve project conventions |
| store_interactionC | Store a conversation interaction for context |
| get_interaction_historyC | Retrieve conversation history |
| cleanup_old_sessionsB | Remove sessions older than specified days |
| get_recent_activityC | Get recent activity for debugging purposes |
| create_taskC | Create a new task in the task tracking system |
| get_tasksC | Retrieve tasks with optional filtering |
| update_taskC | Update an existing task |
| delete_taskB | Delete a task by ID |
| memory_searchA | Full-text search across all session contexts using FTS5 with BM25 ranking |
| search_semanticC | Semantic similarity search (requires @xenova/transformers) |
| search_patternsC | Detect recurring patterns in memory content |
| search_temporalC | Analyze memory patterns over time periods |
| analysis_conflictsC | Detect conflicts between memory entries |
| analysis_memory_mapC | Generate memory map visualization data |
| memory_exportB | Export memories in JSON or Markdown format |
| memory_importB | Import memories from JSON |
| memory_compactB | Compact and optimize database storage |
| memory_tagsC | Manage tags for memory entries |
| task_boardB | Get visual task board grouped by phase, state, or priority |
| task_insightsB | Get task analytics and velocity metrics |
| project_profileC | Manage project profiles for multi-project support |
| get_routing_patternsB | Get learned routing patterns with confidence scores |
| store_routing_patternC | Store a successful routing pattern for cross-workflow learning |
| find_similar_routing_patternsC | Find routing patterns similar to a description |
| store_contexts_batchA | Store multiple session contexts in a single transaction (40-60% faster) |
| track_preferences_batchB | Track multiple preferences in a single transaction (40-60% faster) |
| store_conventions_batchA | Store multiple conventions in a single transaction (40-60% faster) |
| get_tool_manifestB | Return machine-readable tool manifest and required-args schemas |
| store_api_specA | Store and parse OpenAPI/Swagger specification |
| get_api_specA | Retrieve full API specification by ID |
| list_api_specsA | List all stored API specifications with metadata |
| delete_api_specA | Delete API specification and all related endpoints/schemas |
| get_api_endpointsC | Query API endpoints with filters |
| get_api_endpoint_detailA | Get full endpoint details including request/response schemas |
| search_api_endpointsC | Full-text search across API endpoints |
| get_api_schemaA | Get schema definition from API spec |
| analyze_codebase_apisC | Analyze codebase for API endpoints (FeathersJS, Express, Fastify) |
| server_healthB | Check MCP server health and database connectivity |
| record_artifact_readC | Record that a durable on-disk memory artifact was injected or consulted |
| get_artifact_readsB | List recently read durable memory artifacts for feedback and ranking |
| get_autodream_metricsA | List recent autodream runs recorded in the session-memory database |
| server_statsB | Get detailed server statistics and performance metrics |
| store_memoryC | Store a memory with branch awareness, type classification, and importance level |
| get_memoryB | Get memories by topic with branch awareness |
| query_memoryC | Query memory records by keyword |
| update_memoryC | Update an existing memory entry |
| link_memory_to_projectB | Link a memory to a project for operational resurfacing |
| daily_briefingB | Generate proactive daily briefing from operational memory |
| weekly_reviewB | Generate weekly review summary from operational memory |
| stale_work_scanC | Scan for stale work that should be resurfaced |
| assemble_active_contextB | Assemble active context from prompt modules, markdown memory, and SQLite memory |
| search_memoriesB | Full-text search across all memories |
| evolve_memoryB | Add an evolution note to track changes over time |
| get_memory_entitiesC | Get all extracted entities from memories |
| sync_session_startC | Initialize session with branch context and retrieve relevant memories |
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 62 tools
Many tools have overlapping purposes, especially the search and retrieval variants: memory_search, search_memories, search_semantic, search_patterns, search_temporal, get_memory, and query_memory all revolve around looking up stored data with subtle differences. This creates significant ambiguity for an agent trying to pick the right tool.
Tool names mix verb-first and noun-first patterns inconsistently. For example, get_memory and search_memories are verb-first, while memory_search and task_board are noun-first. Verbs also vary (get vs retrieve vs query), and the presence of both memory_search and search_memories for similar actions violates a clear naming schema.
62 tools is far too many for a session-memory server, which typically needs only a handful. The server expands into unrelated domains such as tasks, API specs, routing patterns, server stats, and artifact tracking, making it poorly scoped and overburdened.
For the core session-memory domain, the tool set is quite complete: it covers storing, retrieving, updating, searching, exporting/importing, tagging, compacting, and even analyzing memory/session contexts. There are no obvious dead ends in that main workflow.