Advertise
MCP.so
Sign In
B

Bastra Recall MCP Server

@n0mad-ai

Repository

Basic information

Category

AI & Agents

Transports

stdio

Author

n0mad-ai

Submitted by

Daniel Nevoigt

About Bastra Recall MCP Server

Persistent memory for any AI assistant — Claude, ChatGPT, Cursor — backed by your local Obsidian vault. One daemon, all your tools.

#memory#knowledge-base#obsidian#markdown#search#productivity#personal-knowledge#vault#claude#claude-code#cursor#chatgpt#hooks#rest-api#bm25#vector-search#hybrid-search

Bastra Recall MCP Server config

Add this server to your MCP-compatible client using the configuration below.

{
  "mcpServers": {
    "bastra-recall": {
      "command": "node",
      "args": [
        "/abs/path/to/bastra-recall/packages/daemon/dist/mcp-forwarder.js"
      ],
      "env": {
        "BASTRA_VAULT_PATH": "/abs/path/to/your/vault/memorys"
      }
    }
  }
}

Bastra Recall MCP Server tools

Tool list not available yet

This listing does not currently have a tool list to display. This does not mean the server has no tools; check its documentation for supported capabilities.

We auto-extract tools from the README. The maintainer can list them under a ## Tools heading to populate this section.

Overview

What is Bastra Recall?

Bastra Recall is a persistent memory layer that works across multiple AI surfaces — Claude, ChatGPT, Cursor, and any tool that speaks MCP or HTTP. It runs as a single local daemon serving all AI tools at once, storing memories as plain markdown files with YAML frontmatter in an Obsidian-compatible vault.

How to use Bastra Recall?

Install the daemon locally and point it to a vault folder of markdown files. The server autonomously saves learned information and recalls relevant memories before actions via hooks like PreToolUse and SessionStart (Claude Code). No manual configuration steps beyond setup are described in the README.

Key features of Bastra Recall

  • Single daemon serves every AI tool simultaneously
  • Vault is a folder of plain markdown with YAML frontmatter
  • Hybrid BM25 + vector search with recall_when field weighting
  • Typed memories (lesson, decision, preference, project-fact, workflow)
  • Scope-based isolation per project
  • Wikilinks automatically build a related[] graph

Use cases of Bastra Recall

  • Maintain consistent context across Claude Code sessions
  • Share memory between ChatGPT, Cursor, and Claude Desktop
  • Keep project-specific facts isolated by scope
  • Edit and sync memories via iCloud / Google Drive / Dropbox

Frequently asked questions

1

How does Bastra Recall differ from embed-everything memory approaches?

Bastra Recall uses hybrid BM25 + vector search with the `recall_when` field as the highest-weighted trigger, providing much better recall semantics than approaches that embed everything.

2

Where does Bastra Recall store data?

All memories are stored locally in a user-specified folder of plain markdown files with YAML frontmatter. This vault is compatible with Obsidian and can be synced via iCloud, Google Drive, or Dropbox.

3

What runtime or dependencies are required?

Bastra Recall runs as a single local daemon. The README does not specify a particular language runtime or dependency list, but it communicates over MCP and HTTP.

4

Does Bastra Recall support authentication or bring your own key?

Yes, it is BYOK (bring your own key) and MIT licensed.

5

What transport protocols does Bastra Recall support?

It supports both MCP (Model Context Protocol) and HTTP, allowing integration with any AI surface that speaks either protocol.

Comments

More AI & Agents MCP servers

HeyLead: LinkedIn Outreach & Lead Generation MCP Server

D4umak

HeyLead: AI agent for LinkedIn outreach. Finds the right people, writes in your voice, follows up and handles replies, from Claude Code, Cursor, any MCP client or a web dashboard.

9
L

Legion MCP

faulkj

MCP-native LLM councils for debates, juries, blind panels, voting, refinement, and custom multi-model deliberation.

1

Ranla

Kurtis Tryber

AI growth agent for software businesses. Ranla ships pages that rank on Google, gets your product named in AI answers, replies on Reddit and sends lifecycle email. Ask how growth is going, have Ranla draft the work and a

GoBuy - Product Trust

gobuy-ai

GoBuy gives AI agents marketplace-evidence trust scores. One call returns a 0–100 Evidence Score with observed/missing signals, freshness, and product context. Tools: • check_product_trust — inspect one Amazon, Walmart,

FLORA

FLORA is a remote MCP server available at https://agents.flora.ai. The FLORA MCP lets users generate, organize, and iterate on visual assets — hero images, moodboards, lifestyle shots, promo graphics — directly from a co

Advertise