Our take
Activeregular commits and releases- Merges more pull requests than 98% of the projects we track
- Releases more often than 85% of the projects we track
PLUR is an active, cross-tool memory system for developers who value local control, with a young but regularly maintained project and contributors beyond its leading contributor.
Good fit if
- You use compatible agents and want corrections or conventions to persist across sessions.
- You want memory stored as editable local files, with optional Git-based device or team sync.
- Your environment can run the Node.js tooling and has enough memory for local embedding search, or you are comfortable with keyword-only fallback.
Look elsewhere if
- You need a general codebase index or source-code intelligence, which PLUR explicitly does not provide.
- Your agent environment cannot use PLUR's tools or integrations, and you need automatic memory injection rather than relying on the model to call tools.
- You require a hosted memory API or a temporal knowledge graph instead of a local-first file store.
All health signals
| Last commit | 2026-10-07 (2 days ago) |
|---|---|
| Commits, last 90 days | 100+ |
| Releases, last 12 months | 42 (latest v0.21.5, 2026-10-07) |
| Contributors | 8 (top contributor: 56% of commits) |
| Issues closed, last 90 days | 38+ (typically closed in 1 day) |
| Pull requests merged, last 90 days | 96 (typically merged in 0 days) |
| Project age | 7 months |
Checked on 2026-10-10 with the GitHub API.
Overview
PLUR stores agent-learned corrections, preferences, conventions, and event records in a local memory store. It helps prevent agents from repeating mistakes and lets compatible tools draw on the same knowledge across sessions.
Memory is kept as readable YAML on your machine, with search and recall available through MCP tools and integrations. The project is aimed at developers using coding agents who want portable, inspectable memory rather than a cloud-hosted or tool-specific store.
Key Features
- Save and recall persistent memories called engrams, with project and other scopes.
- Search locally using keyword and embedding-based retrieval, with keyword fallback when embeddings are unavailable.
- Inject relevant memories into agent context and use feedback to tune retrieval over time.
- Capture timestamped episodes and query them as an event timeline.
- Connect multiple agent environments through MCP and dedicated integrations, including Claude Code, Cursor, Codex, OpenCode, and OpenClaw.
- Sync memory across devices or teams using Git, with separate behavior for personal and shared remotes.
- Inspect, edit, and delete the YAML memory files directly; an optional local dashboard provides a read-only view.
Use Cases
- A solo developer wants an agent to remember project conventions and corrections between coding sessions.
- A developer switching between supported coding agents wants the same memory available in each tool.
- A team wants to share selected project knowledge while keeping personal memories on individual machines.
- An operations or support engineer wants a searchable timeline of incidents, fixes, and session events.
What you need
Detected in the repository
- Node.js (from package.json)
- A test suite and automated checks on GitHub Actions
License in plain words
Apache-2.0permissive
- Commercial use: yes
- Modify and redistribute: yes
- You must keep: the license, the NOTICE file and a note of your changes
- Share your changes: no
- Includes an explicit patent grant from the contributors:
A summary, not legal advice: the LICENSE file is what applies.
Getting Started
The README's install path requires Node.js 20 or newer:
npm install -g @plur-ai/cli@latest @plur-ai/mcp@latest
plur init
plur doctor
Restart your editor after setup. See the README for integration-specific instructions and configuration details.
Alternatives
- memoripy: Memoripy is a Python runtime that screens evidence and tracks changing facts with citations, rather than storing readable YAML memory files.
- Memori: Memori captures conversations and execution context as structured persistent memory for applications, rather than using a shared local YAML store.
- EverOS: EverOS stores agent conversations and context in editable Markdown with local retrieval indexes, rather than organizing memory as YAML files.
- ai-memory: ai-memory builds a git-backed Markdown wiki from coding-agent activity and supports handoffs, rather than providing a general YAML memory store.
| Project | Language | License | Stars | Status |
|---|---|---|---|---|
| plur | TypeScript | Apache-2.0 | 298 | Active |
| memoripy | Python | Apache-2.0 | 694 | Active |
| Memori | Python | Other | 17.1k | Active |
| EverOS | Python | Apache-2.0 | 13.4k | Active |
| ai-memory | Rust | MIT | 9.1k | New |
Considerations
PLUR is a young project, though its recent commit and release activity is regular. It has multiple contributors and is not measured as concentrated around one person. Issue handling is brisk, but the repository also has a substantial open-issue backlog, so teams should check whether their specific integration or feature needs are addressed.
Local embedding search requires downloading a model and the README notes a 2 GB RAM minimum; when embeddings are unavailable, search falls back to keyword matching. Team sharing and device sync require careful remote configuration: personal Git remotes can include private-visibility memories, so the README advises using a private remote. The project is Apache-2.0 licensed.
Found this useful?
Share it with someone who would like plur.