Create stream forks referencing source data up to a specified offset.
npx skills add https://github.com/durable-streams/durable-streams --skill forking
Please help me install this Agent Skill. Skill: forking Source: https://github.com/durable-streams/durable-streams/tree/main/packages/client/skills/forking Command: npx skills add https://github.com/durable-streams/durable-streams --skill forking
Forks solve the challenge of reusing and diverging data from an existing stream without duplicating content. They enable independent lifecycles, separate TTLs, and separate deletion of forked data while preserving the source's history up to the chosen offset.
Create a fork of an existing stream by specifying the source and optional offset.
High-intent search queries and answers about installing and using this skill.
Stream forking creates a new stream that references source data up to a specified offset without copying it. This allows independent lifecycle management and separate TTLs while preserving the original stream's history.
When branching from a source stream, forked streams inherit TTL, expiry, and content-type rules. This ensures the fork maintains consistent data expiration and formatting properties derived from the original stream configuration.
Yes, you can read and write to forked streams using a fresh handle without affecting the source. This enables safe experimentation on divergent data variants while the original stream remains completely isolated.
Stream-Forked-From and Stream-Fork-Offset headers define the source stream reference and the specific data offset for branching. They establish the exact boundary up to which the fork inherits source history.
No, forked streams adhere to a server-side deletion lifecycle independent of the source. Deleting a fork only removes the forked variant, leaving the original source stream and its data fully intact.
Use stream forking to create dedicated experimental branches without duplicating content. It enables independent lifecycles and separate TTLs, making it ideal for diverging data analysis while preserving source history.
brainqub3Write and link structured knowledge entries to the kb/ directory.
zapierAutomate read, write, and manage operations in Google Sheets via API v4.
zhixuan312Register large documents once and reference them by ID in mma-* calls.
scooter-lacroixExecute end-to-end tests for the complete data persistence pipeline.
