MCP Memory
Fast Agent Memory Using Google's OKF and SQLite FTS5
Details
- External ID
- 49286073
- Source
- HN
- Company
- —
- Product
- MCP Memory
- Website domain
- github.com
- Launched
- Aug. 13, 2026
- Cohort
- —
- Upvotes
- 70
- Upvotes percentile
- 0.8911290322580645
- Tags
- —
- Fetched at
- Sept. 10, 2026, 5:32 a.m.
- Updated at
- Sept. 10, 2026, 5:32 a.m.
Enrichment
- Theme
- low-level systems and developer tools
- Vertical
- Horizontal
- Function
- Agent / copilot
- Audience
- Developer
- AI stance
- AI-native
- Project type
- Hobby / open-source project
- Normalized one-liner
- memory system for ai agents
- Manually corrected
- False
Could you build this?
Yes This is a Model Context Protocol (MCP) server implementing SQLite FTS5 and Google's OKF for text retrieval, which is standard CRUD and prompt tooling easily generated with AI assistance.
Discussion
20 comments analyzed.
Competitors mentioned: AGENT.md / AGENTS.md, ContextVault, Markbase, Memento, OpenAI Codex and OpenCode
Concerns raised: Claude models perform worse than OpenAI models for tool usage and long-focused work, At hundreds of notes, token consumption becomes prohibitive just to retrieve one thing, Merge conflicts in markdown files depend heavily on instructions and data structure, Opus 5 made remote memory tooling integration more difficult, Not suitable for codebases; better suited for durable facts and cross-project information
Feature requests: Vector embedding and semantic search with pgvector for more accurate retrieval, Git-backed memory with FTS5 as optimization layer for human review capability, Indexing corpus with semantic links (explicit and embedding-based) for faster retrieval, ETag-based validation to avoid merge conflicts, Open source the tool with community help
Competitors
Other products that read as similar to this one — 1178 launches clear the similarity bar, closest 8 shown.
Attention rank: #124 of 1179 (itself plus its competitors, highest first — normalized so YC and Product Hunt are compared fairly).
Launched 287 days after the earliest competitor.
- SQLite Memory · hn · 2026-04-07 · 10 upvotes · similarity 0.79
- A local-first memory store for LLM agents (SQLite) · hn · 2025-12-14 · 48 upvotes · similarity 0.74
- Engrim · hn · 2026-09-07 · 92 upvotes · similarity 0.70
- We made a back end agnostic SDK for Agent Memory · hn · 2026-05-15 · 5 upvotes · similarity 0.65
- Mnemosyne Local hierarchical memory engine for AI agents (MCP Native) · hn · 2026-08-23 · 10 upvotes · similarity 0.64
- How we made web agents 3x faster by building the missing memory layer · hn · 2026-09-30 · 5 upvotes · similarity 0.63
- Hebbs · hn · 2026-03-09 · 6 upvotes · similarity 0.61
- TidesDB · hn · 2025-11-09 · 14 upvotes · similarity 0.60
Other launches for this product
- No other launches for this product.
Same idea, different domain
Nobody's really built a agent / copilot tool for Agriculture yet.