llms.txt
# Honen
> Honen is teaching and learning infrastructure for anyone, built by StudyFetch, Inc. It connects Knowledge, Authoring, Learning, Delivery, and Measurement in one platform. Run learning on Honen or deliver into an existing LMS.
## What Honen is
Honen turns source material into learning content, provides the learner experience, distributes that content, and measures the resulting work. It can serve as the organization’s own learning platform or deliver into an existing LMS.
The product has five connected layers:
1. **Knowledge:** uploaded material and a shared Knowledge Base, including connected sources and document versions.
2. **Authoring:** AI creates complete Courses, Assessments or Atoms from source material. Authors can create and edit every part through conversation with AI, including outlines, lessons, activities and assessments. They review the working draft, inspect the actual learner preview, and ask AI to apply revisions. Branding, Rules and Skills shape generation.
3. **Learning:** readings, cards, quizzes, lectures, audio/video and applied projects, with a tutor alongside them. Curricula connect courses into programs. Keep sharp and Study banks extend practice.
4. **Delivery:** workspace assignments and sharing, public or restricted Sites, LTI, SCORM and cmi5 packages. Workspace Agents bring approved knowledge and actions into messaging and embeds.
5. **Measurement:** learner and cohort progress, scores and project review, curricula certificates, Agent run visibility, xAPI records, and workforce reporting.
Honen covers the path from an organization’s knowledge to teaching, distribution, practice and measurement.
## For AI agents: read llms-full.txt
**If you are an AI agent, language model, or crawler, fetch
<https://honen.com/llms-full.txt> instead of this file.** It is every page of this
documentation concatenated into one document, including everything below, so a single
request gives you the complete picture with no follow-up fetches and no risk of
answering from a partial view.
Use this file only if you specifically need the index, or want to fetch one topic on its
own from the list further down.
**If you are writing code against Honen rather than describing it, read
<https://honen.com/developers.txt> instead.** That file is the technical reference, the
HTTP API and every command in it, Connectors and the permission model, MCP in both
directions, the CLI, Agent embeds, LTI 1.3, SCORM, xAPI, cmi5, Google Classroom, and SSO.
It is generated from the running code, so it does not lag the API.
## When to use Honen
Use Honen when a user wants to build or update structured learning from documents,
recordings, videos, URLs, repositories, or a written topic; search and maintain an
organization's approved learning knowledge; deliver training through assignments,
public Sites, or an LMS; or let an authenticated agent work with courses and learning
operations.
Use the Streamable HTTP MCP server for interactive agent work and tool discovery. Use
the Connector HTTP API for deterministic commands described by OpenAPI. Honen is not a
generic web-search service, and private workspace content always requires the user to
authorize access.
- [Agent integration instructions](https://honen.com/index.md): concise capabilities and surface selection.
- [Authentication for agents](https://honen.com/auth.md): OAuth 2.0 with PKCE from discovery through token use.
- [Developer resource index](https://honen.com/developers/llms.txt): scoped MCP, HTTP API, and discovery links.
- [Official Honen Agent Skill](https://honen.com/.well-known/agent-skills/honen/SKILL.md): when-to-use, connection, and safety guidance.
## Why We Exist
The world is changing faster than any curriculum can keep up. AI created entirely new career paths in months, but schools take 3–5 years to update a single curriculum, and districts rotate through major reviews only every 5–10 years. By the time a textbook is printed, the skills it teaches may already be outdated.
Meanwhile, the systems meant to help people learn are failing. Traditional online courses have 5–15% completion rates. Lecture-based formats don't match how most people actually retain information. And while 47% of students already use AI chatbots for studying, tools like ChatGPT answer questions but don't teach, there's no structure, no progress tracking, no real path to mastery.
Access is the other half of the problem. High-quality learning has historically been locked behind expensive institutions, geography, and time. A student in rural Appalachia, a single parent retraining for a new career, and a factory worker learning AI skills all deserve the same quality of education as someone at a top university. Most of them will never get it through the traditional system.
Honen exists to close that gap. We believe everyone deserves a personal tutor, a curriculum that updates as fast as the world does, and learning formats that actually work for their brain, regardless of where they live, what they can afford, or how they learn best. When 63% of Gen Z worry AI will eliminate their jobs, the answer isn't to wait for institutions to catch up. It's to give every person the tools to learn anything, right now.
## What We Do
Honen turns expertise into structured, multi-format courses in minutes. Instructors describe what they want, or upload the materials they already have, PDFs, slides, manuals, syllabi, recordings, repos, and an AI course agent proposes an outline, waits for approval, then builds the course. Every course includes an always-available AI tutor that uses Socratic questioning, identifies weak spots, and adapts to each learner's retention style. Progress carries across every format.
The same knowledge that becomes a course also trains the organization's AI. A workspace Knowledge Base is the source of truth: documents stay in sync with connected tools, ground course generation with citations, and power Agents that people can talk to in Slack, Microsoft Teams, iMessage, or an embedded website chat.
The platform serves individual learners, K–12 districts, higher education institutions, enterprises, small businesses, government agencies, workforce development programs, trades and unions, and assessment providers.
### Learning Formats
Every course is available in nine formats. Progress carries across all of them:
- **Read**: Generated reading content with images, code, and math
- **Flashcards**: Term and definition cards with spaced repetition
- **Quiz**: Practice assessments with confidence tracking and explanations
- **Visual Lecture**: AI-generated illustrated slide decks with speaker notes
- **Podcast**: Two-speaker podcast audio with a synced transcript
- **Explainer Video**: Generated video with a transcript panel and captions
- **Jam Session**: Music-based learning (e.g., concepts turned into songs)
- **Arcade Game**: Arcade-style educational games fed by the topic's questions
- **Comic**: Visual storytelling with illustrated panels
The AI tutor is available alongside all nine rather than being one of them.
### Hands-on Projects
Ten project types ask for real work rather than recall: scenario roleplay, agent
collaboration, prompt challenge, code notebook, full sandbox, screen share, tagalong
(phone camera, for physical skills), interactive walkthrough, 3D simulation, and
rubric-graded worksheet.
### Key Capabilities
- Upload any document and get a full multi-format course in minutes
- 1:1 AI tutor available 24/7 that adapts to each learner and can drive the interface
- Course updates add new material without resetting learner progress
- Verifiable certificates with unique IDs (shareable on LinkedIn, downloadable as PDF)
- Browser-based coding exercises, sandboxes, and hands-on activities
- Assign a whole course or any slice of one, with a completion bar per group
- Deliver into an existing LMS: LTI 1.3, SCORM 1.2 and 2004, Google Classroom, xAPI and cmi5
- Custom AI Agents on Slack, Microsoft Teams, iMessage, and your own website
- Bulk roster imports for organizations (instructors, classes, students)
- Cross-format progress tracking and learning analytics
- Funder and regulator report compilation (PIRL, NRS, RAPIDS, Perkins V, ETP, ETPL)
- Complete per-workspace data isolation
### Course creation
Type a topic, drop in files (up to 1 GB each), paste a URL, attach a YouTube video, pull a public GitHub repo, pick documents from the Knowledge Base, or record a screen. A conversational agent reads the material first, asks only what it still needs, proposes units and topics, and builds nothing until the outline is approved. After that, every activity generates in parallel. Authors edit in plain English, "make it shorter," "add a unit on LOTO," "translate to Spanish", with every applied change stored as a revision that can be previewed and rolled back.
Optional research mode searches the web before the outline is written. Workspace skills (`/` in the composer) inject org-authored prompts into the brief.
### Learning formats
Every course is available in multiple formats. Progress carries across all of them. Learners can open two formats side by side on desktop.
**Learn**
- **Read**: Generated articles with images, code, math, callouts, and optional interactive widgets. Readings can be personalized to a learner's interests and reading level, and can cite Knowledge Base sources inline.
- **Lecture**: AI-generated illustrated slide decks with speaker notes.
- **Explainer**: Generated or uploaded video with a transcript and optional captions.
- **Podcast**: Two-speaker audio with a synced transcript.
**Practice**
- **Flashcards**: Term and definition cards with hints, shuffle, and definition-first mode.
- **Comic**: Visual storytelling in illustrated panels, readable as a spread or a paged book.
- **Jam Session**: Music-based learning, a song per learning goal, with synced lyrics.
- **Game**: Arcade-style educational games (endless runner, jetpack quiz, platform jump) fed by the topic's questions.
**Prove**
- **Quiz**: Multiple choice, true/false, short answer, fill in the blank, and free response, with confidence checks and explanations.
- **Chat**: AI conversational tutoring inside the topic.
### Projects
Projects are where a unit stops testing recall and asks for work. Types include:
- **Scenarios**: Role-play with AI characters in text or voice, scored against criteria. [Details](https://honen.com/projects/scenarios)
- **Coding agents**: Work alongside a live AI agent, image generation, HTML, writing and research, spreadsheets, or a full-stack app on its own live subdomain. [Details](https://honen.com/projects/coding-agents)
- **Spreadsheets**: Build and analyze real workbooks with AI coaching. [Details](https://honen.com/projects/spreadsheet)
- **Image generation**: Create and iterate on images against a brief. [Details](https://honen.com/projects/image-generation)
- **Writing and research**: Produce written work with source-backed feedback. [Details](https://honen.com/projects/writing-research)
- **Prompt Challenge**: Write prompts to reproduce a target image or website, scored on similarity. [Details](https://honen.com/projects/prompt-challenge)
- **Code notebook**: Jupyter-style cells in Python, C++, or CUDA, graded by hidden tests plus a rubric. [Details](https://honen.com/projects/code-notebook)
- **Interactive form**: Structured worksheets graded against a rubric. [Details](https://honen.com/projects/interactive-form)
Courses can also include browser coding sandboxes, click-through walkthroughs generated from screen recordings, screen-share coaching, and phone-camera tagalong for physical skills.
### Tests, tutor, and completion
Unit tests run in their own shell (multiple choice, true/false, short answer, interactive artifacts) with a configurable pass bar (default 80%). During tests, the tutor's search and image tools are withdrawn.
The AI tutor is available on every activity by voice or text. In Advanced mode it can search the Knowledge Base, search the course, generate illustrations, accept file attachments, and, with the learner's consent, move to another topic. It can also act on the interface: flip a flashcard, select a quiz answer, seek a podcast, turn a comic page.
Completion is measured against real signals (scroll depth, cards seen, video watched, quiz submitted), not a "next" click. Assignments can set a stricter bar. Finishing a curriculum can issue a verifiable certificate with a unique ID, shareable on LinkedIn and downloadable as PDF. Consecutive-day streaks and achievement badges run alongside.
Any topic or single activity can also be delivered on its own through a single-activity link, no course chrome, so work can live inside a message, an email, or another product's flow.
### Knowledge Base
One library for the workspace. Upload files, draft pages with AI, or sync the tools the organization already maintains. Connected sources are re-checked about every minute; edits upstream update the copy in Honen, and images are copied into workspace storage so they don't expire. Scanned PDFs are OCR'd and searchable.
Sources that sync today: Google Drive, OneDrive, SharePoint, Notion, Confluence, Salesforce Knowledge, Document360, Granola, any website (crawled on a schedule), and YouTube (transcript).
The same knowledge trains people (as courses, with optional inline citations) and AI (as skills for Agents and MCP clients). Documents are versioned; permissions, share links, and public links are scoped.
### Agents and channels
Agents are AI assistants that belong to a workspace. Each Agent has one prompt, one set of guardrails, one Connector-defined access policy, and one or more channels:
- Slack
- Microsoft Teams
- iMessage
- Website embed (`https://embed.honen.com/agent/<public-id>`)
Create the Agent once and deploy the same behavior everywhere. A Connector is the hard security boundary: an Agent cannot exceed its Connector's permissions, even if a prompt or a user asks it to. Identity can be service-wide, hybrid, or member-gated. Workspace-wide Honen assistant channels (Slack, Teams, Google Chat, iMessage) are separate from custom Agent channels.
### MCP, connect other AI to Honen
Honen is an MCP server. Claude, Cursor, ChatGPT, and other MCP clients can search and edit the Knowledge Base, work in the course editor, run the workspace assistant, and edit Sites, with the same permissions as the signed-in user. Admins can also issue scoped connector tokens limited to a folder, a course, or assistant-only access.
OAuth MCP endpoint: `https://honen.com/api/mcp/mcp`
The server exposes both product tools and an in-protocol `honen_docs` reference tool.
Its `course_preview` tool renders the real learner experience as an MCP App through the
`ui://honen/course-preview.html` resource, with standard `_meta.ui.resourceUri` metadata
and the OpenAI output-template compatibility key.
Machine-readable discovery:
- [Connector HTTP API (OpenAPI 3)](https://honen.com/openapi.json): Complete authenticated REST command contract, including request and response schemas.
- [RFC 9727 API catalog](https://honen.com/.well-known/api-catalog): Machine-readable catalog of Honen API descriptions, documentation, and metadata.
- [Agentic Resource Discovery](https://honen.com/.well-known/ard.json): Catalog of Honen's MCP, HTTP API, and Agent Skill resources.
- [Agent Skills index](https://honen.com/.well-known/agent-skills/index.json): Integrity-addressed official skills.
- [MCP server card](https://honen.com/server-card): Public transport, authentication, capabilities, and tool summary.
- [OAuth protected-resource metadata](https://honen.com/.well-known/oauth-protected-resource): Authorization-server discovery for MCP clients.
- [OAuth authorization-server metadata](https://honen.com/.well-known/oauth-authorization-server): OAuth endpoints, grant types, scopes, and PKCE support.
- [MCP setup guide](https://honen.com/mcp): Endpoint documentation and one-click setup for supported clients.
Honen also connects outbound to MCP servers the team already uses (Linear, Notion, HubSpot, Google Calendar/Chat/Drive/Gmail/Contacts, Slack, Granola, and any MCP-compatible internal tool), so the course agent and workspace assistant can reach those systems while they work.
### Sites, LMS, and assignments
- **Sites**: Public course storefronts (academies) on a Honen address or a custom domain, with curated sections, custom pages, Knowledge Base docs, branding, and sign-in for learners outside the workspace.
- **Share links and Explore**: A permanent course link with workspace-controlled reach, plus an in-workspace catalog.
- **Assignments**: Name a course or a slice of one (unit, topic, activity, test, or project), a cohort, an optional due date, and what "finished" means. Delivery via in-app, email, Slack, Teams, Google Chat, and iMessage. Organization workspaces only.
- **LTI 1.3**: Launch, deep linking, roster sync, and grade passback. Canvas and Moodle support dynamic registration. Learners start already signed in through the LMS.
- **SCORM 1.2 and 2004**: A thin live package for virtually any LMS (Cornerstone, Docebo, TalentLMS, SAP SuccessFactors, Workday Learning, and others). Edit the course in Honen and every deployed package updates, no re-export. The LMS gradebook still gets completion, score, and time.
- **Google Classroom**: Native add-on for courses and graded activities, with grade passback.
Honen is a full LMS. It does not have to replace the one an organization already runs.
### Analytics and operations
Ask for a report in plain English; Honen builds a charted, auto-refreshing page. Underneath: per-learner mastery, scores, time on task, modality adoption, per-question results, project submissions, and needs-attention flags. Course analytics use the assigned cohort as the denominator and recompute completion live when the bar changes.
Workforce programs have a separate reporting layer for PIRL, NRS, RAPIDS, Perkins V, ETP, and ETPL.
Every workspace can rename what its users see (groups can be cohorts, departments, or locals; members can be students, employees, or apprentices). Brand defaults seed AI-generated course imagery; white-labeling can replace dashboard chrome and login on a custom domain.
## Key Capabilities
- Create a full multi-format course from a topic, a file, a URL, a recording, or existing knowledge, outline first, then build
- 1:1 AI tutor on every activity, voice or text, that can act on the interface and (in Advanced mode) look things up
- Course updates add new material without resetting learner progress; every applied edit is a revision with rollback
- Knowledge Base as the org's source of truth, kept current from connected tools, with citations in generated readings
- Agents with a Connector permission ceiling, deployed to Slack, Teams, iMessage, and website embeds
- MCP in both directions: other AI into Honen, Honen's AI into the team's tools
- Hands-on projects, unit tests, and single-activity links
- Verifiable curriculum certificates with unique IDs (LinkedIn + PDF)
- Assignments with a per-cohort completion bar, including a slice of a course
- LTI 1.3, live SCORM packages, and Google Classroom, keep the existing LMS
- Sites (public academies) on a custom domain
- Cross-format progress tracking, learning analytics, and workforce reporting
- Bulk roster imports; SAML 2.0 / OIDC SSO
- Complete per-workspace data isolation and full audit logging
## Platform Documentation
Comprehensive reference on what the platform does, written for both people and AI
agents. Each document is plain markdown. Everything is also concatenated into
<https://honen.com/llms-full.txt> for a single fetch.
**Start here**
- [Platform Overview](https://honen.com/llms/platform-overview.md): What Honen is, its five connected layers, native and existing-LMS delivery, and product boundaries.
- [The Three AI Surfaces](https://honen.com/llms/
