GitHub hosts HunyuanVideo, Tencent's open-source framework for large-scale video generation models, enabling AI-driven video creation.
LinkLore
FreeStarting price
0Popularity
About LinkLore
LinkLore provides a structured memory system for AI coding agents, enabling persistent project context across sessions and tools. It stores notes in a local SQLite file and automatically reloads them with a brief() call at the start of each session. The system detects code changes and flags contradictions in memory entries to prevent drift. A CLI tool (llre) allows adding, comparing, and managing notes, while an MCP server integrates with compatible agents such as Claude Code, Cursor, Windsurf, and Gemini CLI. Memory can be shared across team members with role-based permissions and optional backend sync. The tool is designed to function as a single shared knowledge base rather than per-agent memory.
Key features
- Local SQLite storage with automatic session reload
- CLI tool for adding and managing notes
- Contradiction detection in memory entries
- MCP server for cross-agent compatibility
- Team sharing with viewer/editor/owner roles
- Audit log for tracking changes
- Code-linked memory that updates with project changes
- Open source components (GitHub, PyPI)
Use cases
- Maintaining project context for AI coding assistants
- Sharing team knowledge across developers
- Tracking project decisions and rationale
Pros
- Persistent project context across sessions and tools
- Automatic contradiction detection in memory entries
- Team sharing with role-based permissions
- CLI and MCP server for integration with multiple agents
- Local-first operation with optional cloud sync
Cons
- Tied to MCP client ecosystem
- Team features require paid plans
- No free tier for team sharing
Frequently asked questions about LinkLore
What is LinkLore and what problem does it solve?
LinkLore is a structured memory system for AI coding agents that provides persistent project context across sessions and tools. It solves the problem of AI agents losing track of project details between sessions by storing notes in a local SQLite file and automatically reloading them with a brief() call.
Which AI coding agents are compatible with LinkLore?
LinkLore is compatible with multiple AI coding agents, including Claude Code, Cursor, Windsurf, and Gemini CLI. It functions as a single shared knowledge base rather than per-agent memory.
How does LinkLore handle contradictions in memory entries?
LinkLore detects code changes and flags contradictions in memory entries to prevent drift. The add() function checks new entries against existing ones and alerts users to potential inconsistencies.
Can LinkLore be used collaboratively by a team?
Yes, LinkLore supports team collaboration with an optional backend sync layer that allows sharing memory across teammates and projects. It includes role-based permissions for viewers, members, and owners.
How do I get started with LinkLore?
To get started, you can install the CLI tool (llre) and use commands like llre add to save notes. The system also provides a brief() function to restore project context automatically at the start of each session.
Is LinkLore stored locally or in the cloud?
LinkLore stores notes in a local SQLite file by default, ensuring fast access and offline functionality. An optional backend sync layer is available for sharing memory across devices and team members.
