joa

joa

Persistent activity journal for AI agents - enables logging and querying decisions, changes, errors, and observations across sessions.

Category
Visit Server

README

joa

CI npm License: MIT

Persistent activity journal for AI agents. Log decisions, file changes, errors, and observations — then query them for context across sessions.

What is joa?

AI agents lose context between sessions. joa gives them a journal — a structured log of what happened and why, queryable by any agent on any platform. Start a session with catchup to see what happened last time. Log meaningful events as you work. Future sessions inherit the full picture.

Works as a CLI tool and as an MCP server for agent platforms like Claude Code, Cursor, Gemini CLI, Codex, Amp, and more.

Install

npm install -g @neethan/joa

Requires Node.js >= 18.

Quick Start

# Log an entry
joa log "Chose PostgreSQL over MongoDB for user data" -c decision -t project:api

# See recent activity
joa catchup

# Search the journal
joa search "PostgreSQL"

# Check journal health
joa status

MCP Server Setup

joa runs as an MCP server so agents can log and query entries directly.

Claude Code (~/.claude.json or .mcp.json):

{
  "mcpServers": {
    "joa": {
      "command": "joa",
      "args": ["mcp", "--agent", "claude-code"]
    }
  }
}

Cursor (~/.cursor/mcp.json or .cursor/mcp.json):

{
  "mcpServers": {
    "joa": {
      "command": "joa",
      "args": ["mcp", "--agent", "cursor"]
    }
  }
}

Or run joa setup for interactive configuration.

<details> <summary>All supported platforms</summary>

Gemini CLI (~/.gemini/settings.json):

{
  "mcpServers": {
    "joa": {
      "command": "joa",
      "args": ["mcp", "--agent", "gemini-cli"]
    }
  }
}

Codex (~/.codex/config.toml):

[mcp_servers.joa]
command = "joa"
args = ["mcp", "--agent", "codex"]

Amp (~/.config/amp/settings.json):

{
  "amp.mcpServers": {
    "joa": {
      "command": "joa",
      "args": ["mcp", "--agent", "amp"]
    }
  }
}

OpenCode (~/.config/opencode/opencode.json):

{
  "mcp": {
    "joa": {
      "type": "local",
      "command": ["joa", "mcp", "--agent", "opencode"]
    }
  }
}

GitHub Copilot (VS Code mcp.json or .vscode/mcp.json):

{
  "servers": {
    "joa": {
      "type": "stdio",
      "command": "joa",
      "args": ["mcp", "--agent", "github-copilot"]
    }
  }
}

Pi (~/.pi/mcp.json):

{
  "mcpServers": {
    "joa": {
      "command": "joa",
      "args": ["mcp", "--agent", "pi"]
    }
  }
}

</details>

CLI Commands

Logging:

joa log <summary>       Log an entry
  -c, --category <cat>  Category (decision, change, observation, error, ...)
  -t, --tag <tag>       Tag (repeatable, e.g. project:api)
  --thread <id|new>     Thread ID or "new" to start one
  --detail <json>       Structured detail as JSON

Querying:

joa query               Query with filters
joa catchup             Recent entries (last 7 days)
joa threads             Active threads summary
joa timeline            Chronological entries
joa decisions           Decision entries
joa search <term>       Full-text search

Maintenance:

joa status              Journal health and stats
joa rebuild             Rebuild SQLite index from JSONL
joa export              Export entries as JSONL to stdout
joa import <file>       Import entries from JSONL (or - for stdin)
joa setup               Configure MCP for agent platforms
joa config get|set      View or update configuration

Run joa <command> --help for detailed usage.

How It Works

joa writes entries to JSONL files (one per day, append-only) as the source of truth, then indexes them in SQLite with FTS5 for fast full-text search and filtered queries.

~/.joa/
  journals/
    2026-02-27.jsonl    # Append-only daily logs
    2026-02-26.jsonl
  journal.db            # SQLite FTS5 index (derived, rebuildable)
  config.yaml           # Optional configuration

If the SQLite index is ever lost or corrupted, joa rebuild reconstructs it from the JSONL files.

Runs on both Node.js (published CLI) and Bun (development).

Contributing

See CONTRIBUTING.md.

License

MIT

Recommended Servers

playwright-mcp

playwright-mcp

A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.

Official
Featured
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.

Official
Featured
Local
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

Official
Featured
Local
TypeScript
VeyraX MCP

VeyraX MCP

Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.

Official
Featured
Local
graphlit-mcp-server

graphlit-mcp-server

The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.

Official
Featured
TypeScript
Kagi MCP Server

Kagi MCP Server

An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

Official
Featured
Python
E2B

E2B

Using MCP to run code via e2b.

Official
Featured
Neon Database

Neon Database

MCP server for interacting with Neon Management API and databases

Official
Featured
Exa Search

Exa Search

A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.

Official
Featured
Qdrant Server

Qdrant Server

This repository is an example of how to create a MCP server for Qdrant, a vector search engine.

Official
Featured