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

Give your agents a head start.

One plan that keeps personal assistants, research agents and coding agents informed, equipped with skills and working overnight.

Your agent, one step of a task
  1. 1Search the library
    intelligence_search "MCP server from an OpenAPI spec"
    openapi-to-mcp-serverSkill, MIT, 6.4 KBsha256 3f9c1a07… Chosen
  2. 2Download the exact version
    provisioning_read expected sha256 3f9c1a07…Checksum matches
  3. 3Put it where the agent reads skills
    step-2-build-server/
    ├── CLAUDE.md
    └── .claude/skills/
    └── openapi-to-mcp-server/

One plan for most of what your agents need.

Compare plans

Feeds

Research. Sources your agents can check.

Freethrough December 31, 2026 (Eastern)

Then $4.99 a month

Includes Agent Feeds

  • Source collections and change feeds
  • Model, benchmark, tool and paper sources
  • JSON Feed, RSS and Markdown for agents
Try Agent Feeds

Paid plans start when you opt in.

Pro

Build. The files each step needs, ready to use.

$29a month

The whole library, 1,560,856 files

Includes Agent Feeds + Skills and agent files

  • Skills, tools, code, hooks and settings
  • Exact versions with checksums, via MCP or API
  • A Baltor key for each tool you connect
Get started

Model usage stays with your provider. Cancel any time from your account page.

Max

Public preview

Keep working. Long jobs run overnight, checked by morning.

$49a month

Supervised runs, local or in the cloud

Includes Agent Feeds + Skills and agent files + Overnight work

  • Supervisors for overnight runs and AFK runs (while you are away from the keyboard)
  • Cloud tools alongside your local ones
  • Limits, checkpoints and a morning report
Set up overnight work

Runs with your own model access. The local tools are free and open source.

Each step gets exactly what it needs.

Your agent searches a library of 1,560,856 skills and agent files, downloads the exact version it chose and puts it in the folder it reads.

Counted October 25, 2026. Identical files count once, and each download names its source, licence and exact version.

Skills
SKILL.md, scripts/, assets/
Instruction files
AGENTS.md, CLAUDE.md
Tools and code
Scripts, functions and utilities
Subagents and commands
Agent and command definitions
Hooks
Checks your agent runs on its own
MCP server settings
The tools a step may call

Built for teams that check the details.

Your model access stays with you

Your agent calls your models through the provider connection you control.

Exact versions

Every download names the selected version and carries a checksum for its files.

You control execution

Your agent runs selected files with the permissions you choose.

Hand off big jobs, one checked step at a time.

Each step names the plan it needs: Feeds, Pro or Max.

See the showcase

Upgrade an app to the next major framework version

Read the breaking changes, pull codemods for your stack, upgrade one tested change at a time.

  1. 1Read the release notes and breaking changesFeeds
  2. 2Pull the migration skill, codemods and test hookPro
  3. 3Apply one change set per step, tests after eachPro
  4. 4Queue the long tail overnight, review in the morningMax

Turn your internal API into an MCP server

Give every agent typed, permissioned access to your own services, from the spec you have.

  1. 1Compare MCP patterns and servicesFeeds
  2. 2Pull a server skill and contract schemasPro
  3. 3Generate tools from your API spec, with testsPro
  4. 4Connect it to the agents you already usePro

Match the right model to every step

Route small steps to small models and keep the large model for the hard part.

  1. 1Read benchmark, cost and latency sourcesFeeds
  2. 2Check which models fit your hardwareFeeds
  3. 3Pull settings and subagents that route each stepPro
  4. 4Run the job overnight within the call limits you setMax

Connect your agent.

One address serves assistants and apps that take connectors, plus Claude Code, Codex, OpenCode, Pi and 17 more tools that connect over MCP.

Open the connection guide
  1. 1Create your accountChoose a plan, then create a Baltor key for the agent you connect.
  2. 2Add one entry to your agentThe address and the name of your key go in the settings your agent already reads.
  3. 3Ask your agentIn plain words. It searches, shows each file’s source and licence, then downloads the exact version.
MCP address
https://baltor.ai/mcp
Transport
Streamable HTTP
Your key
BALTOR_SERVICE_TOKEN, by name

Questions, answered.

Which agents does it work with?

Any AI agent that can read a link or take an MCP connector. Setup guides cover Claude Code, Codex, OpenCode, Pi and the Baltor Harness.

Which models can I use?

The model access you already have: a provider key, Ollama Cloud (https://ollama.com/v1) or an OpenAI-compatible cloud endpoint.

What counts as a download?

One item at the version you selected counts once. Searches, second looks and refused downloads are free.

What does Baltor keep about my work?

Account and download information, plus anything you choose to submit. Your project and model keys stay on your machine.

Start with Feeds, free through 2026.

Add Pro for the whole library, or Max for overnight work, when you need them.