| FMRS | 70 / 100 · B | 75 / 100 · B |
|---|
| Reliability | 9 / 20 | 12 / 20 |
|---|
| Security and permissions | 14 / 20 | 16 / 20 |
|---|
| Maintenance | 18 / 20 | 14 / 20 |
|---|
| Documentation | 17 / 20 | 18 / 20 |
|---|
| Setup experience | 12 / 20 | 15 / 20 |
|---|
| Best for | - Privacy-conscious researchers and developers who want to run LLMs and search fully locally
- Users already running local models via Ollama, LM Studio, or llama.cpp
- People who want to give Claude multi-engine academic search and structured report generation
- Local, STDIO-based integration into Claude Desktop or Claude Code
| - Individuals or teams who prioritise privacy and offline operation and do not want data leaving the machine
- Users who share memory across multiple MCP clients such as Claude Code, Claude Desktop, Hermes Agent, OpenClaw, Cursor, Codex and Gemini CLI
- Scenarios that need encrypted storage, a hash-chained audit log and verifiable vault migration
- Users who want low-latency recall, a bundled embedding model and no extra LLM calls inside the memory layer
|
|---|
| Not for | - Multi-user or network-exposed deployments — this MCP server has no built-in authentication or rate limiting and is documented as local-use only
- Users wanting a zero-config tool — it requires configuring an LLM provider and a search engine, and sometimes deploying SearXNG
- Teams needing real-time multi-user collaboration features
| - Users who need cloud multi-device sync or a hosted team memory service; moving between machines means locking the vault and copying the file
- Users who expect the server itself to run an LLM that extracts facts and decides what to remember; Compartment explicitly keeps no LLM inside and leaves that to the host model
- Users who cannot keep a passphrase safe or who need a recovery path if it is lost, since Compartment never generates a password, seed or recovery phrase
- Deployments that require network transports such as SSE or streamable-http; this server is stdio only and opens no ports
|
|---|
| Required permissions | - Access to a configured LLM (local endpoint or cloud API key)
- Outbound network access to call configured search engines (arXiv, PubMed, SearXNG, Brave, etc.), unless restricted to local documents only
- Local filesystem read access for analyze_documents to work with private documents
- Read/write access to the local encrypted (SQLCipher) database
| - Read and write the vault file under the user's home directory, by default ~/.compartment/memory.vault, and its sibling settings file
- Access the session directory holding the unlock credential (configurable with COMPARTMENT_SESSION_DIR), which also carries the shared embedding process's Unix socket
- Optionally use the macOS keychain (compartment unlock --keychain is an explicit opt-in, and it survives reboots)
- Write or merge an mcpServers entry into each MCP client's own configuration file, taking a byte-exact backup first
- Install a PostToolUse hook for Claude Code that captures memory files the agent writes
- Serve a read-only dashboard on 127.0.0.1 behind a one-time random URL token
|
|---|
| Risks and side effects | - The docs explicitly warn this MCP server is designed for local use only via STDIO transport, has no built-in authentication or rate limiting, and should not be exposed over a network without additional security controls
- Credentials (e.g. API keys) are held in process memory during active sessions, a routine runtime risk the project mitigates with session-scoped credential lifetimes and core dump exclusion
- Research queries are sent to whichever search engines are configured, and cloud LLM providers (OpenAI, Anthropic, Google, etc.) receive query/document content when selected
- Docker's --network host mode can silently fail to expose ports or misroute localhost on Windows/WSL2/Mac
| - A leaked passphrase means the encrypted content can be decrypted; Compartment generates no recovery phrase and cannot recover a lost passphrase
- Enabling the memory_unlock tool places the passphrase in the model's context, which is why it is off by default
- If a 2FA keyfile is lost, for example a USB stick, the vault cannot be opened even with the passphrase
- `export --plaintext` writes the vault as unencrypted JSONL, so that file must be handled carefully
- Memory content can come from untrusted sources; recall wraps memories with a notice that they are stored data and content can be marked quarantined, but the host agent must still treat memory as data to limit prompt-injection risk
- The capture hook and client integration write into the user's own settings files, although the implementation backs up and merges rather than replacing
|
|---|
| Supported clients | Claude Desktop, Claude Code | Claude Desktop, Claude Code, Hermes Agent, OpenClaw, Cursor, VS Code, Codex, Gemini CLI, Cline, Roo Code, Zed, OpenCode, LM Studio, AnythingLLM, BoltAI, Goose, Kiro |
|---|
| Tools | 8 | 14 |
|---|