Your AI tools don't know your product. We fix that.

ContextCompanion creates the shared knowledge base your whole team needs - developers, requirements engineers, and product owners all working with AI that understands your product.

Book an Intro SessionExplore Context Packs

Why context matters in AI development

Great AI output starts with great context.

ContextCompanion captures your entire product - from arc42 architecture docs and ADRs through data models and requirements to team conventions - and turns it into a structured AI context pack your tools can actually use.

Your repo

  • ›technical/architecture/adr-*.md
  • ›methods/working-agreement/definition-of-done.md
  • ›requirements/*.md
  • ›Jira · Confluence · Linearvia MCP
ContextCompanion

AI Context Pack

  • ✓ Architecture & ADRs understood
  • ✓ Working arrangements loaded
  • ✓ Requirements & tickets indexed
  • ✓ External systems connected
  • Your full product context Code, documents and drawings - connected external systems via MCP servers. ContextCompanion ingests the full picture, straight from your repo.
  • Structures it for AI tools. The raw artefacts become a structured AI context pack that Claude, Cursor, Copilot and other AI tools can consume directly.
  • Your whole team benefits. Developers, requirements engineers, and product owners all get structured output so that they can use it for their specific tasks.

Context Packs - Ready-made context packs

Install directly via apm - the agent package manager.

  • $apm install github.com/contextcompanionio/contextpack-adr

Products - Our Products

AI Context Builder

Tool · Early Access

Join the waitlist

The tool that captures your entire product context - arc42 docs, ADRs, team conventions, requirements, and connected systems - and keeps a structured AI Context Pack in sync with your repo.

  • Connects to your repo (GitHub, GitLab, Bitbucket)
  • Reads, structures, and keeps your architecture docs, requirements, and process definitions in sync
  • Integrates external systems via MCP servers
  • Generates and updates your AI Context Pack automatically
Register interest

AI Context Intro Session

90-120 min · Remote

Ideal next step after the Webinar

A focused working session to map your product's existing context artefacts, identify gaps, and lay the foundation for a structured AI Context Pack - tailored to your specific repo, team, and ways of working.

  • Overview of your existing context artefacts
  • Identification of gaps that cause poor AI output today
  • First AI Context Pack draft for your product, built live in the session
  • Prioritised action plan to improve AI output across your whole team
Get in touch

Early Access - Join the waitlist

Be the first to know when AI Context Builder is available. We'll reach out as soon as early access opens.

Who We Are - Built by people who felt the problem.

Patrick Roos

Patrick Roos

Tech Lead · Software Architect · Software Engineer

Patrick has designed and built software systems across complex product domains. After years of watching AI coding tools generate plausible but context-blind output - ignoring architecture decisions, team conventions, and domain knowledge - he co-founded ContextCompanion to fix exactly that gap.

Florian Bosshart

Florian Bosshart

Business Architect · Product Owner · Requirements Engineer

Florian lives at the intersection of business intent and technical execution. His obsession with clarity - precise requirements, well-defined team conventions, decisions that actually stick - makes ContextCompanion pack useful for every role on your team, not just developers.

Contact - Get in touch

Interested in making your organisation AI-ready? Drop us a message and we'll get back to you.