alibaizhanov/mengram

v2.21.0 Breaking

This release includes breaking changes for platform teams planning a safe upgrade.

Published 1mo AI Agents & Assistants

View tool

✓ No known CVEs patched

Read the diff → Tool health → What is this tool? →

✓ No known CVEs patched in this version

Topics

agent-memory ai-agents ai-memory claude-desktop cognitive-architecture cohere

+14 more

cursor-ai episodic-memory knowledge-graph letta-alternative llm-memory mcp-server mem0-alternative model-context-protocol multilingual-embeddings pgvector procedural-memory python llm semantic-search

Summary

AI summary

Agent mode auto‑enables memory extraction when agent_id is present.

Full changelog

What's New

Agent Mode

agent_mode parameter for /v1/add — extract memories from all speakers (user + assistant). Auto-enabled when agent_id is present.

Provenance Tracking

source and metadata parameters on all memory types — track where memories came from (Slack, email, etc.).

CrewAI Integration

MengramMemory drop-in backend for CrewAI agents.

Recall Quality Overhaul

8 search pipeline improvements: stricter relevance floor, dedup, auto-reclassify
Free-form entity types (removed hardcoded 7-type enum)
Speaker disambiguation in extraction
dry_run mode for /v1/add with A/B testable prompt versions

Other

Annual pricing with 20% discount
80% quota warning emails
Growth plan ($49/mo) between Pro and Business
BM25 ranking improvements (ts_rank_cd with normalization)

Bug Fixes

Fix Claude Code hooks silently swallowing quota errors
Fix API key revoke NaN bug
Fix surrogate characters in search cache key
Adaptive min_rrf threshold for search scoring

View diff on GitHub

Weekly OSS security release digest.

The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.

No spam, unsubscribe anytime.

Share this release

Share on X Share on Bluesky

Track alibaizhanov/mengram

Get notified when new releases ship.

About alibaizhanov/mengram

Human-like memory layer for AI agents with semantic, episodic, and procedural memory. Claude Code hooks (auto-save, auto-recall, cognitive profile). 29 MCP tools, knowledge graph, smart triggers, multi-user isolation. Python & JS SDKs.

All releases →