Constellation is a self-hosted, AI-powered knowledge platform from MindWorx. It pulls scattered project knowledge from 17 sources — Slack, GitHub, Jira, Confluence, Google Drive, email, and others — into one indexed workspace, then runs AI skills like conflict detection, gap analysis, and decision logging. Built for engineering teams and technical leads.
A requirement lives in a spec. A decision hides in Slack. A concern shows up in email. A naming convention slips into a GitHub repo. A dependency is buried in meeting notes. A stakeholder assumption sits in a Jira comment.
On their own, none of these pieces seem big enough to matter. But together, they form the reality of the project.
Every project scatters knowledge across tools, conversations, and people. Each fragment is a star — small on its own, powerful when connected.
Requirements that define what needs to be built — often scattered across documents and versions.
Decisions that happen in real-time but get buried under the velocity of daily communication.
Concerns, approvals, and context that live in long threads only a few people ever read.
Naming conventions, patterns, and technical decisions embedded inside files and commit history.
Dependencies, blockers, and context shared verbally — then roughly captured in notes nobody revisits.
Stakeholder assumptions and requirements hiding in comment threads inside issue trackers.
Constellation gathers the stars of your project universe and reveals the shape they make together. Not by replacing the way teams work, but by grounding their work in what already exists.
It shows up in small delays, repeated conversations, conflicting implementation, architectural drift, and the quiet frustration of "I thought we already decided that."
Constellation exists to remove that hidden tax — by creating clarity from the material teams already have.
Most teams are not failing because they lack intelligence.
They are struggling because their knowledge is fragmented.
Common questions from engineering teams evaluating Constellation.
An AI-powered knowledge platform that pulls your project’s scattered information — across Slack, GitHub, Jira, Confluence, Google Drive, email, and 11 other sources — into one indexed workspace and runs AI analysis on top.
Engineering teams and technical leads who lose time to scattered knowledge: decisions buried in Slack, requirements split across docs, conflicts that don’t surface until a feature ships. Teams using AI coding assistants like Claude Code, Cursor, or Copilot also use it to generate optimized context bundles.
It doesn’t replace them — it reads from them. Constellation ingests your existing docs, channels, and repos as-is and adds AI analysis on top: conflict detection, gap analysis, terminology normalization, decision logs, and a Master Document that synthesizes everything with full citations.
Yes — Constellation is self-hosted. Your project knowledge stays on your infrastructure, not ours. The backend is FastAPI + PostgreSQL with pgvector for embeddings; the frontend is Next.js.
Currently 17: file upload, URL, GitHub, Slack, Microsoft Teams, IMAP email, Gmail, Outlook, Jira, Confluence, Google Drive, Azure DevOps, SharePoint, Notion, Linear, manual entry, and Knowledge Base upload.
Yes. The Backlog Generator pushes generated issues to Jira, GitHub Issues, or Azure DevOps. The Master Document can be pushed to Confluence as a new or updated page. Writeback is available on Pro and Enterprise plans.
Actively developed and open for early adopters. Plans are tiered (Basic, Pro, Enterprise) with quota limits and feature gates. To explore access or pricing, contact MindWorx through the website.
Your project is not chaos.
It only looks like chaos because the stars haven't been connected yet.
Constellation helps teams see the real shape of a project — and reveals the structure hiding inside its universe.