Exodocs connects your code repositories to your documentation hub and alerts your team the moment a page drifts from the systems it describes. No more discovering stale docs when it matters most.
Works with GitHub · Notion · Confluence · Slack · Set up in 10 minutes
Exodocs is built around the concept of a code repository and a documentation hub, not specific products. We're starting with the most common and expanding from there.
Using a different tool? Tell us what you need →
Your team writes excellent documentation. Then six months later, the service it describes has been refactored, renamed, or replaced, and nobody updated the page.
Your on-call engineer opens the Auth Service overview. The procedure references a service that was rewritten three months ago. The incident drags on while they reverse-engineer the current behaviour.
The onboarding guide describes a deployment process that changed twice since it was written. New engineers spend weeks building a mental model based on documentation that no longer reflects reality.
A single misleading page at the wrong moment destroys confidence in all documentation. Teams start ignoring their knowledge base entirely and fall back to asking colleagues directly, creating a bottleneck that doesn't scale.
Detect drift as your code changes, fix the docs that are already stale, or generate a fresh set straight from your codebase. Pick the flow that matches where you are.
Add a webhook to any repository. Exodocs receives push events and reads which file paths changed, which is all it needs to detect drift. For richer, more accurate alerts, give Exodocs read-only repo access. Its AI features can then read diffs and sample files to pinpoint exactly what changed and what likely needs updating.
Tell Exodocs which pages in your documentation hub describe which parts of your codebase, manually or with AI-powered mapping suggestions. One page can map to multiple file patterns like app/services/auth/**.
When code that a documentation page describes changes, that page's freshness score drops. When it crosses your threshold, Exodocs sends an alert with the commit, the matched files, and an AI summary of what likely needs updating.
Edit the page in your documentation hub. Exodocs detects the update, resets the freshness score, and clears the alert. The cycle starts fresh. No manual bookkeeping.
Trigger a scan and Exodocs reads every tracked page alongside the current state of the code it maps to, the same inputs that power its freshness scores. No commit required, scan whenever you want a health check.
For each section that no longer matches the code, Exodocs drafts corrected text and explains what changed. Whitespace, config, and unrelated edits are filtered out, so you only see sections that genuinely drifted.
See the current wording next to the proposed replacement. Edit it inline, approve it, or reject it. Nothing reaches your documentation hub without a human approving it first.
Approved edits are written back into the Notion or Confluence page and the freshness score resets to 100. Stale pages get fixed in minutes instead of sitting on a backlog.
No docs yet? Connect a repo and Exodocs reads your file tree and a sample of key files to understand what the codebase actually does.
Generate technical docs written for engineers (service overviews, API references, runbooks, architecture), product docs written for users (feature overviews, user guides, FAQs, release notes), or both at once. Exodocs proposes the right pages for the audience you choose, each ranked by priority.
High-priority pages are drafted for you automatically. Read each one, regenerate it with different emphasis, or dismiss what you don't need.
One click publishes the finished page into Notion or Confluence, already linked to the code it describes, so freshness tracking starts the moment it goes live.
Every tracked documentation page gets a score from 0–100. It decays when matched commits arrive and resets to 100 when the page is edited. Visual and unmissable.
Alerts include the commit message, author, matched files, and an AI-generated summary of what changed, so the page owner can update without reading the diff.
Exodocs analyses your documentation page content and your repository file tree and suggests which code paths the page likely describes. Review and accept in one click.
Not every commit is relevant. Before applying score decay, Exodocs rates how semantically relevant each commit is to the page's content. Whitespace and config tweaks don't penalise your docs.
Plain-English explanations of what changed and which section of the doc likely needs updating, written for the doc owner, not the engineer who wrote the code.
No docs yet? Point Exodocs at a repository and it drafts a prioritised set of pages (services, APIs, runbooks, architecture) for you to review and publish straight to your hub.
With your opt-in, Exodocs posts a drift comment directly on the affected page in Notion or Confluence, so the owner sees what's stale in the doc itself, not just in Slack.
Connect your documentation hub, Notion or Confluence Cloud, and track every page from one dashboard with section-level freshness. One adapter layer means new doc hubs slot in without changing how you work.
Owner / admin / member roles, optional two-factor authentication, and a full audit log of every change, so you always know who connected what and when.
A single dashboard showing the freshness of every critical doc across your team. Catch rot before it causes an incident or embarrasses a new hire.
Get a targeted alert when code you wrote is described in a documentation page that has gone stale, with context on exactly what changed, so you can fix it in minutes.
Procedure docs and architecture pages flagged before incident response. No more discovering mid-incident that a page references a service deprecated six months ago.
Indirect but powerful: the setup guides, architecture docs, and onboarding pages a new hire follows are more likely to reflect reality. Ramp time decreases when documentation can be trusted.
You're connecting tools that touch your codebase and your internal documentation. We take that seriously. Tokens are encrypted at rest with per-organisation isolation, your source code is never retained, optional two-factor auth and a full audit log keep your account tight, and you can turn off every AI feature if you prefer.
Read our full security page →All tokens encrypted at rest. Database dumps reveal only ciphertext.
We never write to your repositories or documentation pages without explicit opt-in.
We persist file paths and commit metadata only. Diffs and sampled files used for AI analysis are deleted on completion, never kept as a persistent store.
Row-level isolation. No org can ever access another's data.
Join the waitlist. We're onboarding design partners now, with free access for 3 months in exchange for weekly feedback.
We'll reply within 24 hours. No spam, ever.