Practical overview
When it fits
Use it when an AI coding assistant needs project-specific memory between sessions and you want that memory to remain readable, local, and version-controlled.
Who it is for
- Developers using AI coding assistants who want project context and lessons to carry across sessions.
- Teams or individual developers who prefer readable Markdown memory stored and versioned locally.
Documented uses
- Capture coding-agent conversations as dated notes and consolidate them into durable project knowledge.
- Retrieve relevant project context and lessons for later coding tasks.
- Preserve plans and progress in tracked files across assistant sessions.
Requirements & limits
The documented setup requires Node.js 22.18 or newer and Git. The default embedding model downloads approximately 219 MB on first use and then runs offline; lexical search can be selected to skip that model.
- Claude Code uses hooks for automatic session behaviors; other MCP clients such as Cursor, Codex, and Claude Desktop follow bundled rules to perform the same steps.
- Saving a proposed lesson requires explicit approval, including when using non-Claude MCP clients.
- The optional hourly schedule consolidates notes and warms embeddings; it is off by default.
- Memory is stored as plain Markdown in a local wiki with Git history rather than in a cloud service.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Install LLM Wiki Memory on macOS or Linux
Set up the local memory repository and schedule hourly background processing.
Before you start
- macOS or Linux
- Node.js 22.18 or newer
- Git
Steps
Clone the repository into ./.llm-wiki-memory/src.
Run ./\.llm-wiki-memory/src/bootstrap.sh.
Run ./\.llm-wiki-memory/src/bootstrap.sh --schedule hourly to enable the recommended hourly schedule.
Expected outcome
The local memory tool is bootstrapped, with hourly background processing scheduled if the final command is used.
Guide limitations
- The hourly schedule is recommended, not enabled by default.
- Without the schedule or a running webapp daemon, embeddings are not warmed ahead of time; a cold note is embedded during its first search.
Compare options for AI coding assistance
These projects document a common task and share a project role. This is a source-based comparison, not an execution benchmark or a guarantee of interchangeability.
| Project | Documented purpose | Platform | Access and requirements |
|---|---|---|---|
| LLM Wiki Memory | Capture coding-agent conversations as dated notes and consolidate them into durable project knowledge. | CLI, Self-hosted | The documented setup requires Node.js 22.18 or newer and Git. The default embedding model downloads approximately 219 MB on first use and then runs offline; lexical search can be selected to skip that model. |
| Continuum — pouyahasanamreji | Store project lessons so coding agents can retrieve them across conversations. | CLI, Self-hosted, Server | The repository documents local Docker Compose and bare-metal development setups. Bare-metal startup requires Node 22.12.0 or later and pnpm 10.33 or later; the Compose path skips these host prerequisites. |
| Haft | Record the problem a project is solving and the options considered. | CLI, Self-hosted |
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 12 October 2026
- Original website title
- ctxr-dev/llm-wiki-memory · Official homepage title · retained in its source language