Session Think MCP
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| SESSION_DIR | No | Storage location for session files | ./.session-think-sessions |
| SESSION_MAX_RETURN | No | Maximum thoughts returned by default | 50 |
| SESSION_NAME_PATTERN | No | Regex pattern for session name validation | ^[a-zA-Z0-9_-]+(:[a-zA-Z0-9_-]+){2,}$ |
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 |
|---|---|
| thinkB | A persistent thinking workspace that preserves reasoning across sessions. IMPORTANT: Always provide a sessionName parameter with format: category:name:subcategory Examples:
If no sessionName is provided, a temporary session will be generated (TEMP:timestamp:random). The sessionName is used to store and retrieve your thoughts. Use consistent naming to maintain context across conversations. |
| list_sessionsC | List all available thinking sessions with metadata. |
| view_sessionB | View the contents of a thinking session. Returns the last N thoughts by default. |
| delete_sessionB | Delete a thinking session permanently. |
| rename_sessionC | Rename an existing session to a new name. |
| search_in_sessionB | Search for thoughts within a specific session by keyword. |
| search_all_sessionsA | Search for sessions containing thoughts matching a keyword. Returns session indicators, not full content. |
| get_session_infoA | Get metadata about a specific session without loading all thoughts. |
| cleanup_sessionsC | Manually clean up old thinking sessions based on age. |
| find_thought_relationshipsC | Search for thoughts that could be related to current reasoning within a session. |
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 10 tools
Each tool has a distinct resource/action: session lifecycle (list/view/info/rename/delete/cleanup), thought capture (think), and search variants (in-session, cross-session, and relationship discovery). Overlaps are minimized by clear distinctions such as metadata vs. content retrieval and keyword search vs. related-thought discovery.
All names use consistent snake_case and most follow a verb_noun pattern (e.g., view_session, search_all_sessions, delete_session). The only notable exception is 'think', which is a bare verb rather than verb_noun, but it is still readable and predictable.
The 10 tools are well-scoped for a persistent session/thought workspace. Each tool earns its place, covering lifecycle, retrieval, search, and cleanup without obvious redundancy.
Core session lifecycle and search/read operations are well covered, including cleanup, metadata, and cross-session search. A minor gap exists around thought-level mutation (no update/delete for individual thoughts) and session export, though thought permanence may be intentional.