Skip to content

Feature/20260202 new features#3

Merged
leduclinh7141 merged 2 commits into
mainfrom
feature/20260202-new-features
Feb 3, 2026
Merged

Feature/20260202 new features#3
leduclinh7141 merged 2 commits into
mainfrom
feature/20260202-new-features

Conversation

@leduclinh7141

Copy link
Copy Markdown
Contributor

No description provided.

Inspired by claude-mem architecture for ~87% token savings.

Layer 1 - memory_search: Returns lightweight index (id, title, tags, score)
Layer 2 - memory_timeline: Returns context around anchor memory
Layer 3 - memory_details: Returns full content for specific IDs

Workflow:
1. search("auth") → get IDs + titles
2. timeline(anchor: "id") → understand context
3. details(ids: ["id1", "id2"]) → fetch only what you need

This prevents fetching full content upfront, saving tokens.
Backward compatible - memory_recall still returns full content.
Inspired by claude-mem architecture for ~87% token savings.

Layer 1 - memory_search: Returns lightweight index (id, title, tags, score)
Layer 2 - memory_timeline: Returns context around anchor memory
Layer 3 - memory_details: Returns full content for specific IDs

Workflow:
1. search("auth") → get IDs + titles
2. timeline(anchor: "id") → understand context
3. details(ids: ["id1", "id2"]) → fetch only what you need

This prevents fetching full content upfront, saving tokens.
Backward compatible - memory_recall still returns full content.
@leduclinh7141 leduclinh7141 merged commit b2c0d9d into main Feb 3, 2026
9 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant