Skip to content

Memori

v3.3.4 Feature

This release adds 3 notable features for engineering teams evaluating rollout.

Published 14d AI Agents & Assistants
✓ No known CVEs patched
Read the diff → Tool health → What is this tool? →

✓ No known CVEs patched in this version

Topics

agent agent-memory ai ai-memory aiagent chatgpt
+11 more
llm long-short-term-memory memori-ai memory memory-management memory-mcp openclaw-memory python state-management stateful typescript

ReleasePort's take

Moderate signal
editorial:auto 14d

Memori v3.3.4 adds memori_compaction for context restoration and LiteLLM client support. Release fixes OpenAI multimodal array parsing, recall filter validation, and refactors Rust core error handling.

Why it matters: New context tool and expanded LLM client options. Deploy on regular schedule; test thoroughly if using OpenAI multimodal features or recall functionality.

Summary

AI summary

Updates feat, test, and LLM across a mixed release.

Changes in this release

Feature Medium

Added `memori_compaction` tool for context restore

Added `memori_compaction` tool for context restore

Source: granite4.1:8b-q6_K@2026-05-20

Confidence: high

 —
Feature Medium

Added LiteLLM as a supported client

Added LiteLLM as a supported client

Source: granite4.1:8b-q6_K@2026-05-20

Confidence: high

 —
Feature Medium

Added pull request template for improved contribution guidelines

Added pull request template for improved contribution guidelines

Source: granite4.1:8b-q6_K@2026-05-20

Confidence: low

 —
Feature Medium

Improved BYODB, Cloud docs and verified them against SDKs

Improved BYODB, Cloud docs and verified them against SDKs

Source: granite4.1:8b-q6_K@2026-05-20

Confidence: low

 —
Feature Medium

Enhanced npm publish workflow with dry run option and verification steps

Enhanced npm publish workflow with dry run option and verification steps

Source: granite4.1:8b-q6_K@2026-05-20

Confidence: low

 —
Dependency Medium

Bumped version for release

Bumped version for release

Source: granite4.1:8b-q6_K@2026-05-20

Confidence: low

 —
Bugfix Medium

Safely parse OpenAI multimodal arrays in query extraction

Safely parse OpenAI multimodal arrays in query extraction

Source: granite4.1:8b-q6_K@2026-05-20

Confidence: high

 —
Bugfix Medium

Restricted recall filters to valid source/signal pairs

Restricted recall filters to valid source/signal pairs

Source: granite4.1:8b-q6_K@2026-05-20

Confidence: high

 —
Bugfix Medium

Fixed widen _request_headers return type

Fixed widen _request_headers return type

Source: granite4.1:8b-q6_K@2026-05-20

Confidence: low

 —
Refactor Medium

Refactored error handling for Rust core retrieval in Memori

Refactored error handling for Rust core retrieval in Memori

Source: granite4.1:8b-q6_K@2026-05-20

Confidence: high

 —
Full changelog

What's Changed

  • Revise README to open all Python SDK sections by default by @jayyao18 in https://github.com/MemoriLabs/Memori/pull/479
  • Revert Trendshift alt text on Memori readme by @jayyao18 in https://github.com/MemoriLabs/Memori/pull/488
  • Merged Python and TypeScript BYODB pages, updated AI models in examples by @jayyao18 in https://github.com/MemoriLabs/Memori/pull/494
  • Remove irrelevant providers from the Install your LLM Provider section by @jayyao18 in https://github.com/MemoriLabs/Memori/pull/496
  • feat: add pull request template for improved contribution guidelines by @devwdave in https://github.com/MemoriLabs/Memori/pull/498
  • Improved the BYODB, Cloud docs and verified them against the Python and TypeScript SDK source code by @jayyao18 in https://github.com/MemoriLabs/Memori/pull/500
  • Fixed widen _request_headers return type by @jayyao18 in https://github.com/MemoriLabs/Memori/pull/502
  • test: add unit tests for recall formatting functions and format_date_created by @krishna28004 in https://github.com/MemoriLabs/Memori/pull/501
  • fixed (LLM) : safely parse OpenAI multimodal arrays in query extraction by @krishna28004 in https://github.com/MemoriLabs/Memori/pull/505
  • fix(skills): restrict recall filters to valid source/signal pairs by @lborro in https://github.com/MemoriLabs/Memori/pull/503
  • Update Memori MCP documentation and configuration for improved tool u… by @devwdave in https://github.com/MemoriLabs/Memori/pull/506
  • Adding compaction instructions by @gordeee in https://github.com/MemoriLabs/Memori/pull/512
  • Compaction brief updates by @gordeee in https://github.com/MemoriLabs/Memori/pull/513
  • Added compaction updates and consistency improvements by @jayyao18 in https://github.com/MemoriLabs/Memori/pull/515
  • feat(openclaw): add memori_compaction tool for context restore by @rpkruse in https://github.com/MemoriLabs/Memori/pull/516
  • Refactor error handling for Rust core retrieval in Memori by @devwdave in https://github.com/MemoriLabs/Memori/pull/517
  • Update Memori MCP documentation for workflow and tool usage by @devwdave in https://github.com/MemoriLabs/Memori/pull/518
  • Enhance npm publish workflow with dry run option and verification steps by @devwdave in https://github.com/MemoriLabs/Memori/pull/520
  • Ts npm release by @devwdave in https://github.com/MemoriLabs/Memori/pull/522
  • feat(llm): add LiteLLM as a supported client by @RheagalFire in https://github.com/MemoriLabs/Memori/pull/461
  • bump version for release by @devwdave in https://github.com/MemoriLabs/Memori/pull/525

New Contributors

  • @krishna28004 made their first contribution in https://github.com/MemoriLabs/Memori/pull/501
  • @gordeee made their first contribution in https://github.com/MemoriLabs/Memori/pull/512
  • @RheagalFire made their first contribution in https://github.com/MemoriLabs/Memori/pull/461

Full Changelog: https://github.com/MemoriLabs/Memori/compare/v3.3.3...v3.3.4

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 Memori

Get notified when new releases ship.

Sign up free

About Memori

Memori is agent-native memory infrastructure. A SQL-native, LLM-agnostic layer that turns agent execution and conversation into structured, persistent state for production systems.

All releases →

Related context

Beta — feedback welcome: [email protected]