Repository navigation
feat(work): add memory-based task recommendations - #4
Merged
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
New Work drafts do not currently use durable memory to surface likely next actions. Generating those suggestions only when the composer opens would also add latency and repeat provider work.
This adds typed task, automation, and Page recommendations with fixed client-owned icons. Selecting a row copies an editable starting message into the current draft and never submits or performs an action automatically. Recommendations use personal memory plus the selected project's memory, or personal memory alone for projectless drafts.
The server persists results by memory, project, model, and prompt version; single-flights concurrent misses; and warms personal plus six recently active project scopes after activation with concurrency capped at two. Cache hits skip note loading and model generation. Memory edits and forgets purge affected derived results, and stale in-flight work cannot restore them. The public read API cannot bypass the cache.
Generation uses the General settings text-generation model, which defaults to Codex gpt-5.6-luna and remains editable. Codex, Claude, and OpenCode use isolated tool-free generation; unsupported providers fail softly with a retry option.
Validation:
Browser evidence was not captured because computer-use permission was not granted for this task.
Built with GPT-5.6 Sol in Codex.