Getting started
Who is ctx| for?
Teams and use cases that benefit from org-scoped context for coding agents.
Engineering organisations
ctx| is aimed at teams that already use AI coding agents (Cursor, Claude Code, Copilot in agent mode, internal tools) and hit the same wall: the model does not know your architecture, your conventions, or what you decided last month unless it is pasted into context every time.
If you:
- Run multiple services or packages in a monorepo,
- Care about consistency (lint rules, ADRs, security patterns),
- Want agents to discover rather than guess APIs and ownership,
…then centralising context in an org-scoped graph plus MCP is the product bet.
Roles
| Role | Typical value |
|---|---|
| Engineering leads / architects | Encode standards once; agents consume them via ctx_advisor and ingestion. |
| Platform / DevEx | One MCP integration for many agents; fewer bespoke RAG pipelines per tool. |
| IC engineers | Faster onboarding to unfamiliar areas of the codebase with search + graph-backed answers. |
| Security-conscious orgs | Self-host option; data stays in infrastructure you control (see Self hosting). |
What ctx| is not trying to be
- Not a replacement for code review or human judgment on risky changes.
- Not automatic sync of every SaaS doc in the world — connections are deliberate (repos, docs you wire in, tools you configure).
- Not magic accuracy: extraction quality depends on repo shape, indexing, and model configuration; always verify critical decisions in code.
If you are unsure whether your workflow fits, get in touch.