AI Agent in Your Repository
Your docs should keep up with your product. The Theneo Agent reviews every pull request, compares it against your entire documentation project, and works out what is now outdated and what is missing, across your API reference and your user guides. Then it drafts the updates for you to approve. (Beta)
Most doc tooling only reacts to API spec diffs. The Theneo Agent reads what a pull request actually changed, a new feature, a behavior change, or a new endpoint, then looks across your whole project to find every page that should change because of it.
Nothing is published without your approval. The agent only drafts changes; you choose what ships.
How it works
1
Connect your repo
Install the Theneo Agent on your GitHub repository and connect it to Slack so you can review without leaving your workflow.
2
It reviews every pull request
When a PR is opened, the agent reads what actually changed, from new features and behavior changes to new or updated endpoints.

3
It posts an analysis in your repo
The agent comments on the pull request with a documentation analysis: which existing sections are now outdated and what new content is missing.
4
It drafts the changes
From that analysis, the agent drafts the actual updates as a draft version inside Theneo, across your API reference and your guides.
5
You review and approve
Open the draft in Theneo, review the proposed changes, and approve to publish. Decline and nothing changes.
Auto or on demand: with auto-drafting enabled, the agent opens the Theneo draft as soon as its analysis is ready. With it off, it posts the analysis first; you trigger drafting by commenting on the pull request when you are ready.
What it catches
- Outdated content: pages that no longer match how your product behaves.
- Missing content: new features, settings, and endpoints that are not documented yet.
- Every type of doc: API reference, guides, concepts, and permissions, not just specs.
Example: for a PR that adds a review-and-approval feature, the agent flags your Publishing Rights, Workspace Configuration, and Access Management pages, then drafts the new settings, roles, and steps into each.
Why it’s different
Docs that never drift
When your product changes, the matching doc updates are proposed right away, so pages never quietly go stale.
Beyond API specs
It reacts to real product changes, not just spec diffs, so your guides and concepts stay current too.
Always human-approved
The agent drafts; you decide. Every change is reviewed before it reaches your readers.
Lives where you work
Triggered from GitHub, reviewed in Slack. It fits the workflow you already have instead of adding another tool.
Prefer a deterministic pipeline? GitHub Actions updates docs on every PR via a workflow file, and the AI Co-pilot writes descriptions inside the editor.
On this page
- AI Agent in Your Repository