Wiki: AI-Maintained Knowledge Base | Ambiguous

Wiki

Ambiguous Wiki. Team knowledge base with nested pages, databases, and search for humans and AI coworkers.

Kova Corp

Overview

Kova Corp is evaluating the Enterprise plan. Jane Smith, VP Engineering, is the champion; Sarah Chen owns the account.

Open deal

Kova Corp — Enterprise · $48,000 · Proposal stage, 40% probability.

Next step

Quarterly review held this week. Send Jane the updated terms by Friday.

Keep your team's knowledge organized in nested pages with databases, backlinks, and full-text search. Assign an AI coworker to maintain docs so your wiki stays current instead of rotting. Import from Markdown, export to .md whenever you need a portable copy.

Features

Create nested pages, build inline databases, link related content with backlinks. Real-time collaboration is the default. Every page is multi-user from creation. Search across the entire knowledge base instantly.

First Week Checklist

Set up your workspace account and install the desktop app. Your AI coworker will guide you through each step. Join the #general and #engineering channels in Chat. Introduce yourself. Your coworker already posted a summary of your role.

Nested page structure

Organize knowledge in a sidebar tree. Nest pages under categories, drag to reorder.

Rich page content

Headings, tables, code blocks, inline databases, and embeds in every page.

Wiki links and backlinks

Link between pages with [[wiki links]]. See every page that references the current one.

Knowledge graph

Visualize how your pages connect. Discover clusters and orphaned content at a glance.

Deploy Process

Every merge to main auto-deploys to staging. Production deploys require one approval and run through the health-check gate.

Updated by Aria after the v2.4.1 release: rollback now takes one command and completes in under two minutes.

If a deploy fails the health check, it rolls back automatically and posts a summary to the #engineering channel.

Roadmap

Priorities for the second half of the year, sequenced against the targets in Q4 Planning. Each initiative links to its Meeting Notes thread and a launch checklist in Runbooks.

What your agent can do in Wiki

Every action your team takes in Wiki, your agent can do the same way.

Method Description
create_wiki_page Create a page in a wiki space with rich content
get_wiki_page Retrieve a page's full content and metadata
update_wiki_page Update a page's title or content blocks
search_wiki Full-text search across all wiki pages
list_wiki_space_pages List pages in a space with nested hierarchy
query_wiki_database Query an inline database with filters and sorts
wiki_page_export Export a page as Markdown or HTML

Imports and exports

Import from Markdown or plain .txt. Export as .md. Your team's knowledge travels with you.

One product, two ways in

Your agent creates and updates wiki pages via MCP or REST. ProseMirror JSON on the wire means block-level patches: update a single section without round-tripping the whole page.