Glacier vs Others
Glacier is a flow-based project management tool with native MCP integration. It's not trying to be everything. Here's an honest comparison with Linear, Notion, and Outline.
In brief
Linear is the best work tracking tool for engineering teams shipping software at pace. Deep GitHub integration, slick UI, great keyboard shortcuts. Built for teams that run sprints.
Notion is the most flexible all-in-one workspace. Databases, docs, wikis, project views — all in one place. Best for teams that want to customise everything.
Outline is the best self-hosted team wiki. Clean writing experience, good permissions, open source. Best for teams that want to own their documentation infrastructure.
Glacier is built for small, AI-augmented teams that work in continuous flow. MCP-first design means Claude can read your board, create cards, and draft docs as part of the same session — without context switching.
Work tracking comparison
| Glacier | Linear | Notion | |
|---|---|---|---|
| Kanban board | ✓ | ✓ | ✓ |
| Column types (queue/active/waiting/done) | ✓ | — | — |
| WIP limits | ✓ | — | — |
| Cycle time tracking | ✓ | — | — |
| Throughput metrics | ✓ | — | — |
| Flight levels (multi-layer) | ✓ | — | — |
| Sprint planning | — | ✓ | ✓ |
| Story points / estimates | — | ✓ | ✓ |
| Roadmaps | — | ✓ | ✓ |
| GitHub integration | ✓ | ✓✓ | — |
Linear's GitHub integration is deeper — two-way sync, auto-close on merge, branch naming. Glacier links cards to issues and PRs and shows status, but doesn't auto-sync.
Documentation comparison
| Glacier | Notion | Outline | |
|---|---|---|---|
| Rich text docs | ✓ | ✓ | ✓ |
| Docs linked to cards | ✓ | — | — |
| Wiki / knowledge base | ✓ | ✓ | ✓ |
| Database views | — | ✓ | — |
| Self-hosted | — | — | ✓ |
| Block-based editor | — | ✓ | ✓ |
Glacier docs serve dual roles: card-linked context (specs, runbooks, decision records attached to work items) and workspace-level knowledge base (onboarding guides, working agreements, team references) with a nested doc tree.
AI integration comparison
| Glacier | Linear | Notion | |
|---|---|---|---|
| MCP server | ✓ | ✓ (limited) | — |
| Read board state | ✓ | partial | — |
| Create/update cards | ✓ | — | — |
| Read linked docs | ✓ | — | — |
| Query flow metrics | ✓ | — | — |
| AI writing assistant | — | — | ✓ |
| AI search | — | — | ✓ |
Linear has an MCP server but it's primarily read-only. Notion has AI features but no MCP server. Glacier's MCP server is the primary interface — read and write, with workspace scoping and flow metric access.
Team fit
Choose Glacier if:
- You're a small team (1–10 people) shipping software continuously
- You use Claude Code or Claude.ai as part of your workflow
- You care about cycle time and flow over sprints and velocity
- You want AI to have real context about your work, not just respond to pasted text
Choose Linear if:
- You run sprints or need roadmap planning
- Deep GitHub integration is a priority (auto-close, cycle sync)
- You have a larger engineering team with specialised roles
Choose Notion if:
- You need a general-purpose workspace (docs + databases + projects)
- Your team does significant content work alongside engineering
- Flexibility and customisation are more important than opinionated structure
Choose Outline if:
- You want a self-hosted team wiki
- Documentation is the primary use case, not work tracking
Where Glacier still has gaps
Being honest:
- No roadmaps. Glacier doesn't have a timeline or roadmap view. If you need to communicate long-term plans visually, use a different tool.
- No sprint planning. If your team is committed to the sprint model, Glacier won't serve you well.
- Limited GitHub depth. Linear's GitHub sync is significantly deeper. Glacier links cards to issues/PRs but doesn't auto-update on merge or branch events.
- Early stage. Glacier is in private beta. Some features that mature tools have — bulk operations, advanced filtering, custom fields — are on the roadmap but not yet shipped.
What's next for Glacier
The roadmap is focused on deepening the things that make Glacier distinct:
- Monte Carlo forecasting — simulate completion dates using historical cycle time data
- Service level expectations — set targets and track adherence automatically
- Aging WIP alerts — surface cards that have been in active columns too long
- Deeper MCP tools — more actions accessible to Claude, including metrics queries
The goal is a tool where Claude Code and Glacier together replace a significant portion of manual project management overhead.