Personal KB vs team KB.

> Everyone's building a personal knowledge base. The team problem is the hard one.

title: Personal KB vs Team KBtype: comparisontldr: personal is solved; teams are the hard part

Obsidian, Notion, Claude, ChatGPT. For one person, it looks clean and works great. Then you try to do the same for a team.

## The easy part

A personal knowledge base is a solved problem.

For one person, the tools are everywhere and they're good. Notes sync, search works, your AI can read them. The personal knowledge base is the TodoMVC of AI apps. Everyone has built one.

## The hard part

What does a team knowledge base need that a personal one doesn't?

The moment more than one person. Or one agent. Shares a knowledge base, a whole iceberg appears below the waterline. This is where most personal-KB tools quietly fall apart.

Personal KB the surface The team layer everything below the surface
Permissions & access control
Conflict resolution when two docs disagree
Onboarding new people (and agents)
Owner accountability & audit trails
Search that holds up at scale
Knowing what's trusted vs. someone's Slack opinion
Single sign-on & provisioning
Keeping it all isolated and private
## The answer

How Basic Memory handles the team layer.

We built the below-the-waterline part. Each team need, and what Basic Memory does about it:

Team needs and how Basic Memory answers them
Team need Basic Memory
Permissions & access controlOwner, editor, and viewer roles, plus per-project visibility (standard, shared, or private).
Conflict resolutionReal-time collaborative editing built on CRDTs (Yjs). Every keystroke syncs, no merge conflicts.
OnboardingNew people and agents read the team's shared knowledge graph on day one. No tribal knowledge handoff.
Accountability & auditAudit logs for account and access activity, plus a new version on every save you can restore.
Search at scaleFull-text, semantic, and knowledge-graph search across every project. Not grep over a folder.
Trusted vs. noiseA curated knowledge graph you own and structure with schemas. Not an opaque pile of chat logs.
SSO & complianceEntra SSO and TOTP for partner and enterprise deployments, with conditional access policies.
Isolation & privacyPer-tenant isolation. Each organization gets its own database and file storage, no shared tenancy.

What we don't do (yet): Basic Memory doesn't auto-ingest from Slack, Jira, or Drive, and it doesn't run automated staleness or content-gap detection. You bring knowledge in and curate it. We'd rather be honest about the edges than overclaim.

## The killer feature

Skills teach every agent how to use it.

A shared knowledge base only works if agents use it consistently. Basic Memory ships Agent Skills. Pre-built instructions that teach Claude and any agent how to capture, recall, curate, and structure knowledge the same way every time. Drop them in and your whole team's agents work to the same playbook.

memory-capturememory-continuememory-curatememory-researchmemory-tasksmemory-schema

Browse the skills →

The team layer, built in.

One shared knowledge graph for your people and their AI agents. with roles, audit, and isolation. Plain Markdown you own.