-
Notifications
You must be signed in to change notification settings - Fork 13k
Ensure saved checkpoints are flexible enough to support long term sessions #21920
Copy link
Copy link
Open
0 / 10 of 1 issue completedOpen
0 / 10 of 1 issue completed
Copy link
Labels
area/agentIssues related to Core Agent, Tools, Memory, Sub-Agents, Hooks, Agent QualityIssues related to Core Agent, Tools, Memory, Sub-Agents, Hooks, Agent Qualitypriority/p2Important but can be addressed in a future release.Important but can be addressed in a future release.workstream-rollupLabel used to tag epics and features that are associated with one of the three primary workstreamsLabel used to tag epics and features that are associated with one of the three primary workstreams🔒 maintainer only⛔ Do not contribute. Internal roadmap item.⛔ Do not contribute. Internal roadmap item.
Description
Metadata
Metadata
Assignees
Labels
area/agentIssues related to Core Agent, Tools, Memory, Sub-Agents, Hooks, Agent QualityIssues related to Core Agent, Tools, Memory, Sub-Agents, Hooks, Agent Qualitypriority/p2Important but can be addressed in a future release.Important but can be addressed in a future release.workstream-rollupLabel used to tag epics and features that are associated with one of the three primary workstreamsLabel used to tag epics and features that are associated with one of the three primary workstreams🔒 maintainer only⛔ Do not contribute. Internal roadmap item.⛔ Do not contribute. Internal roadmap item.
Type
Fields
Give feedbackNo fields configured for issues without a type.
Today, the existing history checkpoints don't capture compression events. We should move away from just storing raw generated content, to saving a higher-level format that supports compression in a lossless way.
This will help stop the bleeding, but very large threads could still become unwieldy. As part of this work, we should come up with a medium term task to fix checkpoints once and for all, i.e. switch to a DB backed solution for history, or consider using multiple JSON files per checkpoint so we can optimize the save / resume accordingly.