memory-manager-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| AGENT_MEMORY_HOME | No | Override the storage directory for agent memory. Defaults to ~/.agent-memory | |
| AGENT_MEMORY_NO_AUTO_SETUP | No | Set to 1 to disable automatic self-registration of the MCP server on first start. | 0 |
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 |
|---|---|
| initialize_project_contextA | Call this FIRST, automatically, at the start of EVERY chat/session — without the user asking. Detects the current project (via git remote, .agent-memory.json or path), auto-registers it, and returns a briefing plus an AGENT PROTOCOL. If unfinished work is detected, the briefing instructs you to ask the user once whether to continue where they left off (yes/no). All memory bookkeeping (save_memory, create_handoff, finish_session) must happen silently in the background — never ask the user to run memory commands. |
| get_project_contextA | Returns the stored compact project context (name, technology, current task, status, summary, last agent). Lighter than initialize_project_context. |
| save_memoryA | Save a curated memory for the current project. Use for decisions, requirements, architecture, tasks, problems, solutions, progress, facts, preferences, constraints and discoveries. Do NOT save raw conversation text — save distilled, useful information. |
| get_memoryA | Retrieve a single memory by id. |
| search_memoryA | Keyword search across memories, tasks, decisions, handoffs, session summaries and project context. Returns ranked results with scores. |
| get_current_taskA | Returns the most relevant open task for the project, plus other open tasks. |
| update_taskA | Create a task (omit taskId) or update an existing one (title, description, status, priority, related files). Statuses: active, in_progress, completed, blocked, abandoned. |
| record_decisionA | Record an important project decision (optionally with rationale and rejected alternatives). Decisions stay relevant for a long time. |
| get_decisionsA | List recorded decisions for the project (newest first). |
| create_handoffA | Create a structured handoff BEFORE ending or pausing work, so the next agent (possibly in another IDE) can continue seamlessly. Include what was completed, what remains, known problems, changed files and the recommended next action. |
| get_latest_handoffB | Retrieve the most recent handoff for the project. |
| start_sessionA | Start tracking an agent working session for the project. Call when beginning work; call finish_session when done. |
| finish_sessionB | Mark a session as finished. Statuses: completed, interrupted, abandoned. Include a short summary of what happened. |
| save_session_digestA | Compress and store a detailed digest of the ENTIRE current conversation (from the first message to now): what was discussed, decided, built, changed, and where work was left off. Call this silently BEFORE ending or pausing any chat. The digest is stored compactly (max 4000 chars) and automatically injected into the next chat's briefing, so a new session understands the previous conversation and can continue seamlessly. Do NOT save raw transcripts — write a distilled, detailed narrative. |
| delete_project_memoryA | PERMANENTLY delete all stored memory for the current project. Requires confirm=true. |
| clear_memoryA | PERMANENTLY delete ALL memory for ALL projects. Requires confirm=true and the phrase "delete everything" in confirmPhrase. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/AbdulqaderAhmed/memory-manager-mcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server