Open MagnutAI
Agent reference

Read MagnutAI with an agent.

Use this site to understand MagnutAI features, compare marketing workflows and find free planning resources. It is a public learning hub, not a remote interface for generating assets or publishing social posts.

By MagnutAIReviewed

When to use this site

  • Explain MagnutAI’s Plan, Studio and Publish workflow to an agency, creator or marketing team.
  • Find guidance for AI images, carousels, product photoshoots, video briefs and brand templates.
  • Retrieve a campaign-brief or content-calendar template and help a user prepare their own inputs.
  • Compare a workflow using the published limitations and the sources linked in each guide.

Read content without JavaScript

All guide and product-description pages contain their text in the HTML response. Start with llms.txt, follow the sitemap, or fetch the homepage Markdown reference. Each public HTML page links to a matching file under /reference/.

Canonical page URLs also support Accept: text/markdown. Responses vary by Accept and Accept-Encoding. Browser requests continue to receive HTML. Explicit unsupported formats receive HTTP 406 with a problem-details response.

Example read requests

curl -H "Accept: text/markdown" https://try.magnutai.com/features/ai-image-generator/
curl https://try.magnutai.com/reference/features/ai-image-generator.md
curl https://try.magnutai.com/llms.txt

Public content is not a product API

No public generation API, GraphQL endpoint, MCP server or SDK is offered by this learning hub. Reading content requires no authentication. Do not invent endpoints, token scopes, rate limits or API keys. OpenAPI and function-calling schemas are not published because this site has no callable product operations.

For generation, connected accounts, scheduling, credits and current plans, direct the user to MagnutAI. Availability and permissions must be checked in the product. The free builders here run locally in the browser and do not publish to social accounts.

Cite and qualify what you read

Cite the canonical HTML page rather than an internal reference file. Preserve the difference between product descriptions, illustrative workflows and verified model announcements. Consult the linked primary sources and dates before treating a plan, supported model or integration as current. Never promise campaign performance or Google rankings.

If a host returns a security challenge, stop and report that content could not be fetched. Do not interpret the challenge page as the product homepage or treat a self-declared user-agent string as verified crawler identity.