slivingdoc
MCP server giving agents one shared, Git-merged notebook stored in S3.
Last verified:
What is slivingdoc?
slivingdoc is a distributed notebook system that enables multiple AI agents and humans to collaborate on shared notes through pull/commit operations backed by S3-compatible storage. It handles concurrent writes, automatic conflict resolution, and maintains an immutable pack-based history without requiring Git or daemons.
slivingdoc pricing
Pricing model: Freemium
slivingdoc pros
- Seamless concurrent access for multiple agents without locks or leases using optimistic concurrency
- Simple interface using ordinary directories and UTF-8 text files, compatible with standard editors and tools
- Automatic conflict resolution via text markers that users edit and resolve naturally
- Works as both MCP tool and CLI; no daemon or hidden sync—files only change on explicit pull/commit
slivingdoc cons
- Requires S3-compatible bucket and AWS credentials—not suitable for offline-only workflows
- Binary files not supported, limiting content to text-based notebooks only
- Requires Node 22+ for setup; manual conflict resolution still needed when changes overlap
Frequently asked questions about slivingdoc
What happens when two agents commit at the same time?
The system uses optimistic concurrency with If-Match headers on S3. One writer succeeds; the other receives a precondition failure, merges the missed changes, and retries with a fresh proposal.
Can humans and agents work on the same notebook?
Yes. Humans can edit files directly in the same directory while agents work. Changes only sync on pull/commit, so both can work without interference until the next operation.
What storage backends are supported?
AWS S3, Tigris, MinIO, or any S3-compatible storage service that supports conditional writes.