Repository navigation
Conversation
…memories mem::consolidate regrouped the same high-importance observations on every run and asked the LLM for new memories. Its only duplicate check was an exact title match against the memory list read at the start of the run, and the LLM rewords titles, so each POST /agentmemory/consolidate stored new copies of the same facts, and two groups in one run could still produce two memories with one title. - Skip a concept group without an LLM call when a current memory already holds all of its top observations in sourceObservationIds. - Evolve the latest memory that shares at least half of those observations when no title matches, so a rewording becomes a new version. - Match only latest memories, and add memories created during the run to the list both checks read. This removes the unused existingTitles set. - Report the number of skipped groups. Signed-off-by: Rex57 <203883624+Rex57@users.noreply.github.com>
|
@Rex57 is attempting to deploy a commit to the rohitg00's projects Team on Vercel. A member of the Team first needs to authorize it. |
Contributor
|
Warning Review limit reachedYou've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository. Next included review available in 34 minutes. View limit detailsLimit details: You’ve used all 10 included reviews currently available. Review configuration: ⚙️ Run configuration
📒 Files selected for processing (2)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
This branch has not been deployed
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.
Summary
Fixes #1486. Also fixes the in-run case reported in #747.
POST /agentmemory/consolidaterunsmem::consolidate, which regroups the same high-importance observations by concept on every run and asks the LLM for up to 10 memories. Nothing recorded that a group had already been consolidated, and the only duplicate check was an exact, case-insensitive title match against the memory list read at the start of the run. The LLM rewords titles between runs, so every run stored new copies of the same facts. On one real store, five manual runs left 80 memories that were 10 topics × 8 copies.Changes in
mem::consolidate:sourceObservationIdsalready contains every top observation of a group, the group is skipped without an LLM call.supersedesit, old oneisLatest: false.existingTitlesset.skipped.The project-scope guard is unchanged; it moved up so both checks share it. The scheduled
consolidate-pipelinedoes not callmem::consolidate; only this REST route does.Checks
test/consolidate-dedup.test.ts(all three fail onmain):skipped: 1)consolidate-project-scope,consolidation-pipelineandconsolidation-defaulttests pass unchangednpm run build,npm run skills:gen,npm run docs:sync(no changes),npm run skills:checknpm test: 2597 passed, 4 skipped