v1.16.2
Gentleman-Programming/engramv1.16.2Jun 13, 2026by github-actions[bot]
Upgrading?
Release notes summary
Introduces memory lifecycle states (active/needs_review) and review functionality, including MCP tools and HTTP routes.
Key Highlights
- Adds observation lifecycle state derived from `review_after`
- Adds `mem_review` for listing and marking reviewed memories
- Pins critical observations in context to prevent burying
New Features
- Active/needs_review lifecycle states
- MCP `mem_review` tool
- HTTP review routes
- Critical observation pinning
- TUI lifecycle state display
Full Release Notes
## Highlights - Adds observation lifecycle state derived from `review_after`. - Adds `mem_review` for listing memories that need review and marking one reviewed. - Adds Pi-native review support through the Engram HTTP server for `gentle-engram`. - Pins critical observations in context so important memory is harder to bury. ## Changes ### Memory lifecycle - Surface computed observation state as `active` or `needs_review`. - Add store-level review queries and local-only review reset behavior. - Expose lifecycle metadata through MCP save/search responses. ### MCP, HTTP, and Pi - Add MCP `mem_review` for direct agent integrations. - Add HTTP review routes used by `gentle-engram` Pi-native tools. - Add compact Pi chrome for review list and mark-reviewed flows. ### TUI and docs - Show lifecycle state in the TUI list/detail views. - Document `mem_review`, lifecycle metadata, HTTP routes, and Pi setup behavior. ## Install ```bash brew upgrade engram ``` ## Compatibility notes - `mark_reviewed` resets the local review cycle only; lifecycle review timestamps are not synced yet. - Pi users should pair this with `gentle-engram@0.1.8` for native `mem_review` support.
Related Repositories
ollama/ollama★ 184,509
Get up and running with Kimi, GLM, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.
fatedier/frp★ 110,102
A fast reverse proxy to help you expose a local server behind a NAT or firewall to the internet.
infiniflow/ragflow★ 92,781
RAGFlow is a leading open-source Retrieval-Augmented Generation (RAG) engine that fuses cutting-edge RAG with Agent capabilities to create a superior context layer for LLMs
gohugoio/hugo★ 90,215
The world’s fastest framework for building websites.
junegunn/fzf★ 83,750
:cherry_blossom: A command-line fuzzy finder