rememberLABSCO SUMMARY
It remembers why the workflow changed, not just that it did.
Every memory tool stores what happened. The interesting move here is the procedure record: version three of "deploy" arrives carrying the assumption that turned out false — the build container had enough memory — and a precondition that travels with it. An agent picking it up does not have to re-derive the two mistakes that produced it. The rest is a competent hosted memory service; this is the part that is hard to copy.
What it is
A hosted memory service reachable over MCP, plus Python and JavaScript SDKs and a CLI. What separates it from a fact store is the third memory type: procedures carry a version, and reporting a failure produces a new version with the violated assumption and a derived precondition attached, so an agent loading the latest one knows what to check first.
What you get
- By the README's count, 30 tools for memory management over MCP, usable from Claude Desktop, Cursor, Codex, Windsurf, Cline or a hosted Claude agent
- Semantic memory for facts and preferences, episodic memory for events and decisions with their outcomes, and procedural memory for workflows
- Procedures that evolve: report an outcome and a new version is produced carrying
violated_assumption,preconditions, and runningsuccess_count/fail_count - Automatic failure detection — describing a failed deploy in conversation links the episode to the procedure and evolves it without an explicit feedback call
- A synthesised answer with citations that link each claim back to the fact it came from, rather than a raw list of matches
- A cognitive profile generated from all memories in one call, ready to paste into a system prompt
- Retrieval across 23 languages: ask in Russian, Chinese, Spanish or Japanese and get an answer in the same language, cited back to facts stored in their original one
- Multi-user isolation on a single API key, so each
user_idsees only its own memories and profile - Claude Code hooks that load the profile at session start, recall on every prompt and save after responses — surviving
/clearand auto-compaction - History import from ChatGPT exports, Obsidian vaults or plain markdown, with secrets redacted on your machine before upload
Requirements
A Mengram API key, in the om- form, free from mengram.io. server.json publishes a streamable-http remote at https://mengram.io/mcp that takes your key as a bearer token in an Authorization header; the local path is pip install mengram-ai (2.30.0) and running the mengram command with server and --cloud, with MENGRAM_API_KEY set. mengram try runs entirely on your machine with no account if you want to see what it would store first. Self-hosting against Ollama is supported but the extraction prompt is around 4,000 tokens, so the project asks for a model of 8B parameters or more with an 8K or larger context window. Licensed Apache 2.0.
Setup effort
One command plus a key — pip install mengram-ai, then supply credentials
USE CASES
01
Stopping an agent repeating a failure it already hit
Report the outcome when a workflow fails; the procedure version bumps with the failed assumption and a precondition attached, and the next run checks it first.
02
Carrying context through a Claude Code compaction
Install the hooks — the session-start hook fires after /clear and after auto-compaction and re-injects your profile, so a lossy summary does not lose the memory.
03
Personalising any model in one call
Generate the cognitive profile and drop it into the system prompt of whatever LLM you are using.
04
Answering from memory with sources
Ask a question and get a written answer whose citations point at the individual stored facts, so a wrong answer is traceable to a wrong memory.
05
Serving many end users from one key
Pass a user_id on every call; search and profile generation stay scoped to that user.
30 tools
Listed in project code.
remember_textrecallasksearchtimelinevault_statsrun_agentsget_insightslist_proceduresprocedure_feedbackprocedure_historyget_entitydelete_entitylist_episodessearch_allget_graphget_triggersget_feedarchive_factmerge_entitiesreflectdismiss_triggerfix_entity_typelist_memoriesget_reflectionsdedupcheckpointcontext_forgenerate_rules_fileBEFORE YOU INSTALL
What to know first
Tools you get
30 tools registered in the project's code.
Requires
- mengram on PATH
- MENGRAM_API_KEY in your environment
INSTALLATION
Connect your client
Server setup
pip
TERMINAL
pip install mengram-aiHosted endpoint
ENDPOINT URL
https://mengram.io/mcpAdd this hosted endpoint to your MCP client using the credentials listed under Requirements.
MCP Clients
Claude Desktop, Cursor, Other MCP clients
JSON
{
"mcpServers": {
"mengram": {
"command": "mengram",
"args": [
"server",
"--cloud"
],
"env": {
"MENGRAM_API_KEY": "<YOUR_MENGRAM_API_KEY>"
}
}
}
}Claude Code — Plugin
CLAUDE CODE — PLUGIN · TERMINAL
claude plugin marketplace add alibaizhanov/mengram
claude plugin install mengram@mengramClaude Managed Agents
Add Mengram to your agent definition. Then store your Mengram API key in a vault as a static_bearer credential for https://mengram.io/mcp, and pass the vault when creating a session.
CLAUDE MANAGED AGENTS · JSON
{
"mcp_servers": [
{
"type": "url",
"name": "mengram",
"url": "https://mengram.io/mcp"
}
]
}Any MCP client — Hosted Endpoint
ANY MCP CLIENT — HOSTED ENDPOINT · JSON
{
"mcpServers": {
"mengram": {
"url": "https://mengram.io/mcp",
"headers": {
"Authorization": "Bearer <YOUR_MENGRAM_API_KEY>"
}
}
}
}The hosted endpoint answers 401 without a key: <YOUR_MENGRAM_API_KEY> in the Authorization header above is your Mengram API key (free at https://mengram.io).
