Documentation that grows with your code.

Context Central reads every change to your code, gathers the story behind it from the tools your team already uses, and keeps onboarding guides, product docs and API references accurate - automatically.

app.contextcentral · Developer Onboarding Watching your code…
Code change PR #1284 · merged

auth: require 2FA for admin sign-in

12export async function signIn(user) {
13- if (user.isAdmin) return grant(user)
13+ if (user.isAdmin) {
14+ await verifyTwoFactor(user)
15+ }
16 return grant(user)
17}
SEC-212 · “Admins need 2FA”
Developer Onboarding › Signing in

How admins sign in

Admins sign in with their work email and password, then land on the dashboard.

Source: auth/signIn.ts:13 · SEC-212 PA Review requested

The problem

Code changes daily.
Docs change when someone remembers.

Every merged change makes some page a little more wrong. Nobody notices until a new hire follows outdated steps, a customer reads a stale guide, or support answers from last quarter’s version.

  • Trust quietly decays

    Once people find one wrong page, they stop reading all of them.

  • Engineers become librarians

    Hours go into rewriting docs instead of building product.

Code vs. documentation, over a quarter

Your code Your docs

Week 1Week 6Week 12 the “drift” ✓ always in sync

Docs only change when someone finds time. The gap keeps growing.Docs update alongside every change — reviewed, approved and published.

How it works

From roots to canopy.

Like a healthy tree, good documentation grows from strong roots.

  1. 01 · Roots

    Code context - the source of truth

    Code is the only thing that’s always true about how your product works. Context Central connects to your git provider and maps your codebase into a knowledge graph: every service, function, API and how they depend on each other.

    AI agents watch that graph. When a change is merged, they know exactly which parts of the system moved — and which pages describe them.

  2. 02 · Trunk

    Business context - the “why”

    Code tells you what changed. Tickets, specs and planning docs tell you why. Context Central links each change to the conversations behind it, so docs explain intent — not just mechanics.

  3. 03 · Branches

    External context - the bigger picture

    API specs, READMEs and design files fill in the details people actually look for: endpoints, setup steps, screens and flows. Only what’s relevant to the change is pulled in.

  4. 04 · Canopy

    Your inference layer - docs, your way

    You decide what good documentation looks like: the audience, structure, tone and rules for each doc. Context Central uses that inference layer to write and maintain every page — then routes updates for approval before anything is published.

    Onboarding guidesProduct docsAPI referencesAnything you define

Every change, end to end

What happens when code ships.

No one has to remember to update the docs. Here’s the path a single merged change takes.

Code merged

A developer merges a change, like any other day.

Impact found

The code graph pinpoints which docs are now out of date.

Context gathered

Linked tickets, specs and designs explain the why.

Update drafted

Agents rewrite only the affected sections, in your style.

Reviewed

The right owner approves, edits or sends it back.

↺ or back to redraft

Published

Everyone sees the accurate version, with sources attached.

Connectors

Plugs into the tools you already use.

No migration, no new habits. Connect your sources once and Context Central keeps listening — so knowledge scattered across a dozen tools ends up in one accurate place.

CodeGitHub · GitLab · Bitbucket
Work trackingJira · GitHub Issues
DocumentsGoogle Docs · Microsoft Office · READMEs
Specs & designOpenAPI · Figma

…and more connectors growing every month.

Docs you get

One source of truth. Every audience.

The same living context becomes different documents for different readers — each one written the way that audience needs it.

Governance

Automated, never unsupervised.

AI does the heavy lifting. People stay in charge of what gets published and who gets to see it.

Access control

Roles at the organisation and project level decide who can see, edit and approve each document.

Organisation

  • SPOwnerFull control of the organisation
  • NKOrganisation AdminManages every project & its members

Project · Payments API

  • PJProject AdminRuns the project, assigns roles
  • RMEditorWrites and edits docs
  • TLViewerReads published docs

RM Extra permissions for Editor granted by an admin

Approval workflows

Every AI-drafted change waits for the right person to sign off. Nothing goes live by surprise.

  1. Agent drafted update to Signing in2m ago
  2. Review requested from Project Admin2m ago
  3. Changes requested: “mention backup codes”1m ago
  4. Agent revised the draft1m ago
  5. Approved by Project Adminjust now
  6. Published to Developer Onboardingjust now

Who can approve

OwnerOrganisation AdminProject Admin+ anyone granted “Approve updates”

Who it’s for

Everyone who depends on knowing how things work.

Engineering leaders

Faster onboarding and fewer “how does this work?” interruptions.

Developers

Ship code. The docs follow without a separate chore.

Product teams

Product docs that match what actually shipped this week.

Support & success

Answer customers confidently from pages you can trust.

FAQ

Questions, answered.

What does “continuous documentation” mean?

Instead of writing docs once and letting them age, documentation is updated continuously as your product changes. Each time code is merged, Context Central works out what that means for your docs and drafts the update for review.

Why is code treated as the source of truth?

Tickets get closed, specs get outdated and meeting notes get lost — but the code is what actually runs. Context Central starts from the code (its “code context”) and adds business and external context on top, so documentation never contradicts how the product really behaves.

What is the inference layer?

It’s your set of instructions for each document: who it’s for, how it should be structured, what tone to use and what to include or leave out. You provide it once; Context Central follows it every time it writes or updates that doc.

Will AI publish changes without anyone checking?

Only if you want it to. Approval workflows let you require sign-off from an admin, or from any member who has been given approval rights, before an update goes live.

Do my engineers need to change how they work?

No. Developers keep using their usual git provider, tickets and docs. Context Central connects to those tools and works in the background.

Who controls access to our code and documents?

You do. You choose which repositories and sources to connect, and role-based access control decides who can view, edit or approve each organisation, project and document.

Plant it once.
Watch your docs grow.

Connect your code and tools in minutes. See your first always-accurate document the same day.