Have an AI agent? Let it explore Baltor for you. Any assistant works, personal or coding.

Connect your agent
What your agent reads

The request it copies

Explore baltor.ai for me and report back. Start with https://baltor.ai/llms.txt, then read https://baltor.ai/.well-known/agent-card.json and https://baltor.ai/.well-known/mcp.json. Tell me which plan (Feeds at $4.99/mo, Pro at $29/mo, or Max at $49/mo) fits my work, what you would connect, and the setup steps at https://baltor.ai/connect. Stay read-only: read public pages and agent files, and check with me before any sign-in, purchase or change. If a step needs a key, name it (BALTOR_SERVICE_TOKEN) and I will add it myself.

Discovery files on baltor.ai

Read-only discovery is open. Keep BALTOR_SERVICE_TOKEN in your own environment.

Baltor
Sign in Get started

Getting set up

Create an account, add your connection and make a first download. Keep using your own tools and model access.

Updated October 25, 2026

Before you start

  • A Baltor account with Pro for the library. Pro is Agent Feeds + Skills and agent files, $29 a month. Agent Feeds alone is free through December 31, 2026 (Eastern). Compare plans

  • Your harness, installed on the machine that runs the work. Your harness is the program that runs your coding agent, such as Claude Code, Codex, OpenCode or Pi.

  • A model you can reach. A provider key, Ollama Cloud (https://ollama.com/v1) or an OpenAI-compatible cloud endpoint. Model access stays with you.

Create your account

Start with just your email address. Baltor sends the confirmation link itself, from accounts@mail.baltor.ai, with the subject Confirm your Baltor account. The link opens a page where you choose a password of 12 to 72 characters.

Creating an account keeps your plan choice open. Subscribe from your account page when you are ready; card details stay with Stripe.

Get startedSign in

Choose your connection

Every harness reaches Baltor at one address, https://baltor.ai/mcp, over the Model Context Protocol (MCP) with streamable HTTP. Connect with OAuth where your harness offers it, or with a Baltor key.

Connect with OAuth, key-free

Harnesses that use OAuth open a Baltor page that names the app, its return destination and the access it asks for. Check them, then allow access. The connection stays within the scopes you allow: search, downloads, usage or billing.

Use a Baltor key

Create one key for each tool or device on Keys and connections. Baltor shows each key once, so put it straight into your environment. This prompt hides the key while you paste it, and your shell history keeps only the command.

Bash
read -rsp "Baltor service token: " BALTOR_SERVICE_TOKEN
export BALTOR_SERVICE_TOKEN

The configuration holds the variable name; your key stays in your environment. Set it in the shell that starts your harness.

Add Baltor to your agent

Merge one entry into the file your harness already reads, beside your other servers.

Claude Code.mcp.json
{
"mcpServers": {
"baltor": {
"type": "http",
"url": "https://baltor.ai/mcp",
"headers": {
"Authorization": "Bearer ${BALTOR_SERVICE_TOKEN}"
}
}
}
}

Reads .claude/skills/ and CLAUDE.md. Claude Code quickstart

OpenCodeopencode.json
{
"$schema": "https://opencode.ai/config.json",
"mcp": {
"baltor": {
"type": "remote",
"url": "https://baltor.ai/mcp",
"headers": {
"Authorization": "Bearer {env:BALTOR_SERVICE_TOKEN}"
}
}
}
}

Reads .opencode/skills/ and AGENTS.md. OpenCode quickstart

Codex~/.codex/config.toml
[mcp_servers.baltor]
url = "https://baltor.ai/mcp"
bearer_token_env_var = "BALTOR_SERVICE_TOKEN"

Reads .agents/skills/ and AGENTS.md. Codex quickstart

Pi.pi/extensions/baltor.ts
curl -fsSL --create-dirs \
-o .pi/extensions/baltor.ts \
https://baltor.ai/assets/pi/baltor.ts

Start Pi in this folder with the key set. Reads .pi/skills/. Pi quickstart

The Baltor Harness quickstart covers the harness that runs overnight work. 17 more tools connect at the same address: Cursor, Gemini CLI, GitHub Copilot, Windsurf, Cline, Roo Code, Kilo Code, Aider, Goose, Amp, Kiro, Zed, Qwen Code, Crush, Factory Droid, Hermes Agent and OpenClaw.

A ready-to-run worker

One downloadable container includes OpenCode, the Baltor harness, Python, Node.js and Git. Your project and model access stay with you.

Run Baltor on your own computer

Make your first download

Ask your agent in plain words. It searches, shows you the source and licence, then downloads the exact version it chose.

Give your agent this prompt

Search Baltor for a CSV validation tool. Show its source, licence and requirements before downloading it.

What you see

  • Results with their source, licence and version, before anything downloads.
  • The file in the folder your harness reads, with a matching SHA-256 checksum.
  • One line in your usage record: the item, its checksum and the time.

Search in the WorkspaceSearching and retrieving

Check the connection

Ask the service first. Its capabilities record names the protocol versions, the transport, the search modes and the download limits.

Bash
# Ask the service what it serves
curl https://baltor.ai/api/v1/capabilities

On Keys and connections, Test service connection checks metadata only and lists the tools your connection can use.

intelligence_searchprovisioning_manifestprovisioning_readpublic_good_files

Where each agent reads files

AgentReads skills fromInstruction file
Claude Code.claude/skills/CLAUDE.md
Codex.agents/skills/AGENTS.md
OpenCode.opencode/skills/AGENTS.md
Pi.pi/skills/Named in the Pi quickstart
Baltor’s own agentThe task file it is givenThe task file lists the files each step needs

Remove a connection

Revoke the key on Keys and connections. Revocation takes effect on the next request, and files you downloaded stay on your computer. Then clear the variable and remove the baltor entry from the configuration.

Bash
# After you close your harness
unset BALTOR_SERVICE_TOKEN

Need a hand? Status, the connection checklist and the operator's email address are on one page.

Get support