Pick a focused bundle
Start with a domain, workflow, category, policy area, or playbook your AI needs to understand.
Education for AI-ready knowledge
Better knowledge makes better AI.
Give assistants, agents, and automation tools structured, cited, reusable knowledge instead of rebuilding context from scratch.
The problem
AI tools keep asking for context because they do not know your work yet.
Most assistants can write, summarise, and plan, but they still need the right background before their output is useful. That background is often scattered across docs, wikis, tickets, spreadsheets, policies, and team memory.
Knowledge Bundles package that material into structured, source-aware context that a person can inspect and an AI tool can reuse.
How it works
From scattered context to reusable AI knowledge.
Give it to your AI tool
Use the bundle as reusable context for agents, assistants, coding tools, research workflows, or team automations.
Check the sources
Inspect the markdown, metadata, citations, and licence notes before relying on the output.
Reuse and improve it
Keep the same structured context available across repeatable work instead of rebuilding prompts from scratch.
Use cases
Choose the kind of AI work you want to improve.
AI builders
Give agents reusable context, source provenance, and clearer task boundaries before they start work.
Build with better contextBusiness operators
Package procedures, handovers, policies, and service workflows into context assistants can reuse.
Improve operationsCompliance teams
Start AI workflows from traceable sources, licence notes, review prompts, and clearly bounded use cases.
Work from sourcesMarketing teams
Stop rebuilding the same audience, category, positioning, and campaign context in every AI tool.
Reuse the briefEducation teams
Turn curriculum notes, examples, rubrics, and learner support material into AI-ready context.
Package learningCreators
Turn useful expertise into validated bundles with marketplace distribution and an 80% creator revenue share.
Package expertiseWhat bundles can contain
Specific knowledge beats generic prompting.
A bundle should be narrow enough to be useful, clear enough to review, and practical enough to plug into real assistant work.
Concept pages that explain the important thingsSource links and citation notes for claimsMetadata that helps tools identify the topic and use caseLicence and permitted-use notes where they matterExamples, prompts, or workflow guidance for practical useIndex and log files that make the bundle easier to browse and maintain
Bundle examples
Start with a narrow job your AI needs to understand.
Agent workflow references with source notesMarketing category packs for repeatable briefsCompliance guides with citations and limitsOperations playbooks for team handoversEducation packs for learning supportAustralian context bundles for local workflows
Trust layer
Useful AI knowledge should be readable, sourced, portable, and reviewable.
Open Knowledge Format gives Knowledge Bundles a simple structure: markdown files, metadata, links, and citations. That means the knowledge can be read by people and consumed by agents without a proprietary runtime.
The marketplace adds the commercial layer: discovery, samples, requests, creator submissions, review signals, and a path to buy ready-made bundles.
Learn about the formatReadable
Bundles are built from ordinary files and markdown, so people can inspect the knowledge instead of trusting a black box.
Source-aware
Good bundles show citations, provenance, and licence notes so teams can check where important claims came from.
Portable
OKF gives bundles a simple structure that can move across tools, teams, repos, and AI workflows.
Reviewable
Marketplace review focuses on structure, metadata, conformance, and whether the bundle is clear enough to use responsibly.
Comparison
A Knowledge Bundle is more than another place to save prompts.
Prompt library
A prompt library stores instructions. A Knowledge Bundle stores reusable context the prompt can work from.
Team wiki
A wiki is written mainly for people. A bundle keeps the material readable while adding structure for AI tools.
RAG source folder
A source folder can be useful, but a bundle adds metadata, citations, licence notes, and intended use cases.
FAQ
Common questions before you browse.
Do I need to understand OKF to use a bundle?
No. OKF is the structure underneath. Buyers can start with the bundle topic, examples, sources, and use case.
Can I use bundles with different AI tools?
That is the goal. Bundles are designed around portable files and readable context rather than one closed app.
Are bundles guaranteed to be correct?
No. Good bundles make sources, limits, and review notes easier to inspect, but important work still needs human judgement.
What if the marketplace does not have the bundle I need?
Use the request path. Early demand signals help prioritise new categories, first-party bundles, and creator opportunities.
Ready to give your AI better context?