CLI and MCP tools for AI agents

Build courses with AI agents.
Review before you publish.

Give your source material to an external AI agent. It writes course YAML with Graspful tools. Review the lessons and answers, then import a draft and confirm publication.

bun add -g @graspful/cliGet started Free

Works with your tools

An agent with MCP or terminal access can use Graspful tools. The llms.txt reference supplies documentation to other assistants.

Claude Code

MCP tools through Claude Code

Cursor

MCP integration in Composer

OpenAI Codex

CLI tool use with MCP

ChatGPT

Browse llms.txt for context

Gemini

Browse llms.txt for context

Windsurf

MCP integration in Cascade

VS Code Copilot

MCP support in agent mode

Two YAMLs. One product.

A course YAML defines the learning content. A brand YAML configures the product. Import both, and you have a live business.

1. Course YAML

Define your knowledge graph — sections, concepts, prerequisites, knowledge points, and practice problems.

aws-saa-c03.yaml

2. Brand YAML

Configure the landing page, pricing, theme preset, and SEO. This becomes a standalone product site.

aws-prep-brand.yaml

3. Import & Launch

Import the reviewed course and brand files, then confirm course publication and domain verification. Billing requires separate setup.

MCP Tools

MCP tools cover scaffolding, content stubs, validation, review, and import. Browser authentication creates the credentials for publishing.

graspful_scaffold_course

Create a course skeleton for your agent to fill from source material

graspful_fill_concept

Add knowledge point and problem stubs for your agent to author

graspful_review_course

Run 10 automated checks, including problem depth and publication readiness

graspful_validate

Quick schema validation against the course YAML spec

graspful_import_course

Import a course YAML to the platform, optionally publishing immediately (auth required)

graspful_create_brand

Generate a brand YAML with landing page, theme, pricing, and SEO

How agents build courses

Authenticate for import and publish

Install the CLI, then run graspful register when you are ready to import. For MCP, configure GRASPFUL_API_KEY in your agent. Local scaffolding and review work without an account.

Start from a source document

Feed the agent an exam guide, textbook, or documentation. It scaffolds the knowledge graph — sections, concepts, and prerequisite edges.

Fill concepts incrementally

The agent fills one concept at a time — generating knowledge points with progressive difficulty and at least 3 practice problems each.

Review catches quality issues

The review gate checks content readiness, question variants, teaching alignment, and structure. A passing result still needs source and answer-key review.

Import and publish

Import a course draft, request publication, and confirm published: true in the result. Import brand settings separately. Paid subscriptions are not available yet.

Start with local authoring

Local authoring and review are free. Paid subscriptions are not available yet. We will publish plan and payout details when billing is ready.

See current availability

Start with one course draft.

Install the CLI and give your source document to your agent. Author, review, import, then confirm publication.

bun add -g @graspful/cliGet started