pomazanbohdan/memory-mcp-1file

v0.8.2 Feature

This release adds 3 notable features for engineering teams evaluating rollout.

Published 3mo MCP Data & Storage

View tool

✓ No known CVEs patched

Read the diff → Tool health → What is this tool? →

✓ No known CVEs patched in this version

Summary

AI summary

Added minScore support to filter weak matches in search_memory.

Full changelog

v0.8.2 — What’s Updated

Search quality improvements

Added minScore support to search_memory (vector/bm25) to filter out weak matches.
Improved memory result deduplication (content_hash + normalized-content fallback).
Reduced vector-search noise with adaptive score-floor normalization.

Indexing and graph pipeline

Optimized symbol_relation construction:
- moved unresolved symbol fallback resolution to a batch strategy,
- reduced per-reference DB lookups,
- improved candidate selection priority: same file -> same directory -> any.
Added extra relation-stage metrics in logs for diagnostics.

Logging and diagnostics

Raised default log level to debug in code.
Improved fresh quality script (scripts/search_quality_fresh.sh) with:
- post-completion stability wait,
- strict final check for project_info.status == completed.

Validation and testing

Executed full fresh-cycle validation (fresh volume + full indexing).
Verified incremental behavior for add / modify / delete scenarios.
Ran search tool checks across memory/code/symbol/graph flows on completed index state.

Full Changelog: https://github.com/pomazanbohdan/memory-mcp-1file/compare/v0.8.1...v0.8.2

View diff on GitHub

Weekly OSS security release digest.

The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.

No spam, unsubscribe anytime.

Share this release

Share on X Share on Bluesky

Track pomazanbohdan/memory-mcp-1file

Get notified when new releases ship.

About pomazanbohdan/memory-mcp-1file

A self-contained Memory server with single-binary architecture (embedded DB & models, no dependencies). Provides persistent semantic and graph-based memory for AI agents.

All releases →